Build Advice New build questions ?

[Richard1234]

Well, then you need to Google it some more. Or perhaps pay for expedited shipping. Your call.

at the outer level I have tried Google and its a jungle where I cant find it,

possibly if I go to the checkout on a particular one it may have faster delivery.

I am going later to a "DIY" shop in a satellite town, to see what they have, before trying further on the internet.

I find the information given on the unopened products and the stuff supplied with items a bit deficient, eg the Rolson doesnt say that it is 4mm. And the one I bought doesnt say it is 4mm, the manual refers to 4mm and 6.35mm, and the picture in the deficient manual is different from the actual product, and the product was factory sealed, very difficult to open without damaging the box.

Paranoid much?

Well, if you were to download the firmware file from 3rd party site, with questionable source, then yes, there are grounds to be paranoid. File could contain rootkits.

But downloading it directly from MoBo manufacturer's official webpage? It really doesn't get any more official/legit than that. Other than you contacting MSI directly via customer support and then customer support person sending you the firmware file directly. Most likely they refer you to the official download site on MoBo Support page.

not paranoid! but you have misinterpreted what I said!

what I am saying, is you download an update from the manufacturer webpage, and it is fully legit.

but we dont know what the contents are, where I am not saying it contains malware, but that its a black box.

and this is where people "brick" their BIOS, as they were dabbling with something they didnt understand and which was quite different from what they imagined.

Spoiler: this part is unexpanded as it is heading off topic

if you looked inside it would just look like some random hexadecimal, 979CA3BB2FE3310867C.......

in a literal sense we know what is in the file, but not in an interpretative sense.

potentially someone could try to disassemble the bytes, but that might not work if its for an undocumented chip.

as compared to say Wordpad, where its also some random hex, but where we can deduce what it is doing, it reads the bytes from a text file, and displays them in a window. the specifics of how it does that we dont know, but just by reading the manual and using the program we could write our own version which would function identically.

with a big budget we could create something that does the same as the mobo also! as it just has to implement the collection of sockets, CPU, memory, USB, PSU etc. but the updates wouldnt work as those relate to hidden deficiencies of the manufactered one. eg the russians created concordski, their own version of the Concorde.


but with the BIOS update, we cannot write our own version, because we have no idea what it is doing.

the max level of publicly knowable is open source, where you can see the source code, modify this if you want, rebuild to make the used file.

a lower level of publicly knowable is eg say Wordpad, where we know exactly what the software does, from the user manual, and could write our own version.

but if we are given a program, and told: run this program and it will "improve" your system, then what it does isnt publicly available, "improve" is indeterminate information. There is a spectrum of circumstances for some software we are given, from outright publicly available, namely open source, to outright unavailable, where its just an "update" for some subsystem. in the middle is software where we know exactly what the software does, but we dont know how, eg say Internet Explorer or Wordpad. which are both about working with publicly documented formats, eg HTML and related file formats for Internet Explorer.

a lot of updates make no difference whatsoever for most people.

with big companies, maintenance is often used for industrial espionage. eg a firm's photocopier needs fixing each week, and the guy who fixes it walks away with a module which has recorded all images.

 

[Richard1234]

Well, then you need to Google it some more. Or perhaps pay for expedited shipping. Your call.

as I prefer to not assume anything, I went to the checkout on that link, and there is no expedited delivery, earliest is 6th March, see screenshot, so I wont complete that purchase for now, but might later if I cant find any other option.

 

[Aeacus]

at the outer level I have tried Google and its a jungle where I cant find it,

Well, i can only give you links where to buy one, but i can't tell how long the shipping takes.

Before, i actually found one more;
amazon UK: https://www.amazon.co.uk/Adapter-6-35mm-Driver-Magnetic-holder-2pcs/dp/B093ZQV5JY/

Even found 4mm to 4mm extender (to better reach into restricted space),
amazon UK: https://www.amazon.co.uk/Adapter-Extendable-Magnetic-Electric-Screwdriver/dp/B0CB1JCS7K/

But overall, 4mm HEX is quite rare. Most are 6.35mm HEX. Might want to visit horologist workshop and ask from where they bought their tools (i think they should use 4mm HEX bits as well). Or PC repair shop in that matter. They also should use torque screwdriver and 4mm HEX bits.

but we dont know what the contents are, where I am not saying it contains malware, but that its a black box.

We do not have to know what is inside the firmware file nor have we any right to know it.

but with the BIOS update, we cannot write our own version, because we have no idea what it is doing.

There are intellectual property laws, whereby MoBo manufacturer isn't obligated to disclose what is inside the firmware file, used to update the workings of their MoBo (aka BIOS update).
Now, if you want to know what is inside the firmware file, design and build your own MoBo. Nobody is stopping you. And create your very own custom firmware to it, to run as you desire. But as long as you are using someone else's product, you are bound to what OEM dictates.

as compared to say Wordpad, where its also some random hex, but where we can deduce what it is doing, it reads the bytes from a text file, and displays them in a window. the specifics of how it does that we dont know, but just by reading the manual and using the program we could write our own version which would function identically.

Wordpad is a poor excuse of open source software, since it isn't open source. While it is free (you don't have to pay for it), the code that makes it run isn't open source. Sure, you can illegally hack the program files open and decode the contents, to learn how the program works, but that doesn't make the software open source.
In similar sense, nobody isn't stopping you to hack the firmware file and trying to make sense how it works (if you even can comprehend it, that is).

Everything that is to do with Micro$oft and bundled with Windows (namely Micro$oft software), isn't open source.

There are open source software out there and if you like to free access to the source code + customization of it, then feel free to use those software. Almost all GNU/Linux distros (except few that you need to pay for). Regarding text editors: OpenOffice or LibreOffice. Video playback: VLC Player. Web browser: Mozilla Firefox. Image editing: GIMP. Etc etc.
Most of the software that i use with my PC is open source (the ones i listed are also the ones i'm using).

eg the russians created concordski, their own version of the Concorde.

For one, it isn't called "concordski". At best, that is a slang word. What it is actually named, is: Tupolev Tu-144. And Russians didn't "create" it. What they did, were using their spies to steal the building plans from Brits/French and made a copycat version of the Concorde.

Same is with Boeing B-52 Stratofortress, where Russian spies stole the building plans of it as well. In early designs, B-52 was supposed to have turboprob engines, so, Russians stole those plans and created their own copycat version of it: Tupolev Tu-95. But Americans changed the plans before production and put turbojet engines on it instead. Hence why B-52 now has turbojet engines but Tu-95 has turboprop engines.

 

[Richard1234]

Well, i can only give you links where to buy one, but i can't tell how long the shipping takes.

Before, i actually found one more;
amazon UK: https://www.amazon.co.uk/Adapter-6-35mm-Driver-Magnetic-holder-2pcs/dp/B093ZQV5JY/

Even found 4mm to 4mm extender (to better reach into restricted space),
amazon UK: https://www.amazon.co.uk/Adapter-Extendable-Magnetic-Electric-Screwdriver/dp/B0CB1JCS7K/

But overall, 4mm HEX is quite rare. Most are 6.35mm HEX. Might want to visit horologist workshop and ask from where they bought their tools (i think they should use 4mm HEX bits as well). Or PC repair shop in that matter. They also should use torque screwdriver and 4mm HEX bits.

I went to the shop in the nearby town, they didnt have anything, then I tried a small hardware shop they sent me to their building supplies shop, the guy then sent me to screwfix, after much study the staff at screwfix said they dont do it, but that the standard drill bits are 6.35mm which is a round about way of saying 1/4"

I have now decided to purchase a 6.35mm torque screwdriver instead, as I can then just use standard bits, and some of the supermarkets sell comprehensive kits by reputable firms at cheap prices, and the torque screwdriver should arrive as quickly as the 4mm one did,

this one in fact is digital and is torque both clockwise and anticlockwise:

https://cpc.farnell.com/sealey/sts103/1-4-digital-torque-driver-0-05/dp/TL17413

and the price is 48.89 versus some 27 for the 4mm to 6.35mm converter. better to pay an extra 11 quid than to wait approx 2 weeks!

I will enquire if I can get a refund on the other one, but probably cannot in which case the money wasted is collateral damage!

We do not have to know what is inside the firmware file nor have we any right to know it.


There are intellectual property laws, whereby MoBo manufacturer isn't obligated to disclose what is inside the firmware file, used to update the workings of their MoBo (aka BIOS update).
Now, if you want to know what is inside the firmware file, design and build your own MoBo. Nobody is stopping you. And create your very own custom firmware to it, to run as you desire. But as long as you are using someone else's product, you are bound to what OEM dictates.


Wordpad is a poor excuse of open source software, since it isn't open source.

I never said it was open source! I said it is essentially openly specified, ie the manual, the menus etc. so you can basically clone it from the outside without hacking it in any way. where you could externally clone it to create a program which does the same stuff, but any updates wouldnt work. as it would only be externally identical.

Spoiler: unexpanded to prevent thread proliferation as it is a bit off topic,

Linux is an involved example of this kind of thing, where it does the same stuff as Unix, various subtleties are different, and you couldnt run Unix software directly on it, but should be able to rebuild such to it. its in a general sense a "reimplementation" of Unix. as they years go by, it is developing in its own direction, and probably now is better than Unix.

the idea of Linux is recompilation reimplementation, where the Unix program wont work, but if you rebuild the identical source code on Linux it now will work. essentially Linux is source code compatible with Unix, but not binary compatible. In fact GNU takes the idea further, to be able to write source code which rebuilds to any OS at all, for Windows it is via an environment called Cygwin. last time I tried Cygwin was probably before 2010, but if I remember rightly you run some software on the Windows command prompt, and now its just like Linux.

GNU today is different from what it originally was. originally I think it was just to create open source programs for Unix.

AMD originally was a reimplementation of Intel. with the Amiga, people have created an emulator, where you can run the same software, called "Amiga Forever" by Cloanto, but to run piracy protected floppy disk software, you need a hardware bridge from a PC, called catweasel as you cannot run such from a normal PC floppy drive. (a cat and a weasel are 2 totally different animals!)

you have 2 different things: specification and implementation. the specification of Wordpad or any user software is open, the implementation is closed. the specification is the external meaning, the implementation is the hidden realisation of that meaning.

eg HTML is a specification, and Internet Explorer, Firefox, Chrome all implement that same specification.

While it is free (you don't have to pay for it), the code that makes it run isn't open source. Sure, you can illegally hack the program files open and decode the contents, to learn how the program works, but that doesn't make the software open source.

there's no point trying to see how Wordpad works, as it is a very standard work!

Spoiler: unexpanded as a bit off topic

text editors are very standard works, with the Amiga 500 we had Microemacs, which is a great editor, which is based on GNU Emacs. a nice feature is you can split the screen to simultaneously edit several different parts of the same text file, or edit several different files.

In similar sense, nobody isn't stopping you to hack the firmware file and trying to make sense how it works (if you even can comprehend it, that is).

if it can run without CPU, it may not be possible, you'd have to hack the hardware also.

Everything that is to do with Micro$oft and bundled with Windows (namely Micro$oft software), isn't open source.

it isnt open source, but is de facto open specified. anything with an instruction manual which fully specifies what something does, is essentially openly specified. eg "loads and allows you to crop jpegs" is an open specification.

Spoiler: unexpanded as verging off topic

Basically something is openly specified if the public information by the manufacturer enables you to create something which does the same stuff. The BIOS updates are thus not openly specified.

AmigaOS 1.x was unusual in that they openly specified the implementation! with AmigaOS 2.0 onwards they began separating the specification from the implementation. But this specification of the implementation is what makes AmigaOS 1.x very popular with programmers, as you can hack into the OS in a way which isnt viable with Linux or Windows. And it means people who have programmed AmigaOS 1.x generally have a much better understanding of operating systems than people who have only used Unix, Linux, Windows, Apple. I think today Apple is in any case just a front end of Unix, and Android a front end of Linux.

Electronic Arts in fact began with the Amiga. if you went into games stores, eg the Virgin Megastore on Oxford Street, central London, the computer games zone was virtually all Amiga, with just a small area for other computers. With the Amiga 1000, Electronic Arts did some really great non game software for the Amiga, Deluxe Paint (1985) and Deluxe Music Construction Set (1986). It was lightyears ahead of PCs of that time! Also the Amiga 1000 had pre-emptive multitasking with windows, icons, mouse, menus in 1985, and I think with just 256K memory. It took PCs 10 years to catch up with Windows 95. PCs overtook the Amiga for games in the early 1990s BEFORE Windows 95, entirely because their graphics hardware became more advanced. as games players arent interested in the OS, just in the graphics hardware.

even today, Windows hasnt fully overtaken what the Amiga 1000 could do in 1985!

There are open source software out there and if you like to free access to the source code + customization of it, then feel free to use those software. Almost all GNU/Linux distros (except few that you need to pay for). Regarding text editors: OpenOffice or LibreOffice. Video playback: VLC Player. Web browser: Mozilla Firefox. Image editing: GIMP. Etc etc.
Most of the software that i use with my PC is open source (the ones i listed are also the ones i'm using).


For one, it isn't called "concordski". At best, that is a slang word. What it is actually named, is: Tupolev Tu-144.

but it will be forever remembered as concordski! Tupolev Tu-144 is an instantly forgettable "name", concordski is a name you will NEVER forget, not even in 20 years time!

"concordski" is like "obamacare", "reaganomics", "stalinism", "thatcherism" its a nickname by the press,

And Russians didn't "create" it. What they did, were using their spies to steal the building plans from Brits/French and made a copycat version of the Concorde.

Spoiler: expand for off topic ideas on this

quite possibly, or it was a sham, which had the appearance of Concorde, and the alleged crash was a cover story to prevent people calling the bluff!

either way, it was an imitation and concordski is quite an apt name, as it gets to the heart of the matter that it was an imitation of concorde, where they have russianised "concorde" to "concordski".

Concorde in fact was developed within a few miles of where I live, at Filton, Bristol. Today that area has a Rolls Royce facility and an Airbus facility, where Britain I think makes the wings for Airbus, and Boeing recently have set up an office block in the same district.

Concorde ultimately lasted a long era as a supersonic passenger craft for eg London to New York, whereas concordski vanished as a passenger craft after 7 months!

Same is with Boeing B-52 Stratofortress, where Russian spies stole the building plans of it as well. In early designs, B-52 was supposed to have turboprob engines, so, Russians stole those plans and created their own copycat version of it: Tupolev Tu-95. But Americans changed the plans before production and put turbojet engines on it instead. Hence why B-52 now has turbojet engines but Tu-95 has turboprop engines.

Spoiler: expand for very off topic comments, probably best continued via private messages

the clever MO is to include some bluff in case of spies, which is a form of counter espionage.

Britain spent £52.8 billion on defence in 2023! which is the highest in europe other than Russia, and higher than Japan.

the main problem the british military had historically is they got infiltrated by Russia, eg Kim Philby. The fall of the soviet union was because of Reagan's starwars program, which bankrupted Russia. Before Obama the US military I think was 8x as big as any other military, even China and Russia, Obama was trying to reduce that to 4x! But did you know that the UK's NHS employs more people than the US military! before WW2, Britain was the most powerful military, based on its navy. Russia's navy has problematic access to the oceans. Britain, Portugal, France control the atlantic! and the US controls the pacific and the atlantic.

 

[Richard1234]

anyway, the bitter lesson I have learnt is that if you want to buy a torque screwdriver, only get a 6.35mm = 1/4" one, as everything is geared to 1/4" ones. And eg much easier to get 1/4" to 4mm converters than vice versa. You need to note whether the torque range covers the torque you need, as there is major variance, many dont cover the torque I need.

the 1/4" ones also are a lot cheaper, where the one I bought is some £48, whereas 4mm ones are easily £90. thus 1/4" plus converter to 4mm should be a lot cheaper, than a 4mm one!

but the scene is such a jungle, that I could only discover this by making a lot of mistakes

 

[Richard1234]

I have printed out the mobo manual, and one advantage is I can annotate this,

studying this, various questions:

on p23 of the mobo manual, where for some inexplicable reason the physical page number and absolute page number are the same! I am sure earlier these were 2 different numbers!

button 1 is the "clear CMOS" button, this wont work then if you brick up the BIOS?

it says "Power off your computer. Press and hold the Clear CMOS button for about 5-10 seconds to reset BIOS to default values".

I dont know about today's mobos, but in my era, mobos had a watch battery on them, is this entirely for the clock, or is it for storing the "early startup" settings, eg boot drive order?

the BIOS updates, these are recorded to special memory that persists even when powered down like with a flash drive?

ie if the watch battery runs out, the BIOS isnt affected?



with p23 of the mobo manual,

2 + 10 + 12 are USB 3 C sockets, which look the same in the diagram, but the info says:

socket 2 is USB 3.2 Gen 2x2 20Gbps (from ASM3241) and this conflicts with PCI_E3,
socket 12 is USB 3.2 Gen 2x2 20 Gps (from X670 chipset),
socket 10 is USB 3.2 Gen 2 10Gps (from CPU) "support DisplayPort 1.4 with HBR3"

now I am a bit confused, I thought the graphics would need the highest speed, but this appears to suggest we need socket 10 for graphics, which presumably is half the speed of the other 2?

also with the other 2, as socket 2 conflicts with PCI_E3, in the sense that the one is at the cost of the other, presumably this means one would use socket 12 before socket 2?

also who or what is ASM3241? it looks like a password!

you said in the other topic that buses are from either the chipset or the CPU, but here we have apparently a 3rd entity, the ASM3241.


the block of sockets 3, 11, 10 all say "from CPU", the next block #4 says from chipset,

so should I be using the 3 + 11 + 10 block before the chipset ones, or it makes no difference?

could you recommend a long USB 3 C cable which will handle the max video speeds of this mobo?

also any recommendation for a short cable to convert the other end of the USB3 C cable to DisplayPort for my monitor?

as that would be female USB3 C, I dont know if I would need a gender converter USB3 cable as well?

or if I need a long male to female USB 3 cable?

because I have to go from female USB3 C socket to female DisplayPort socket,

where I thus need either:

USB3 C male to female, long cable, then male USB3 C to male DisplayPort short cable,
OR
USB3 C male to male, long cable, then female USB3 C to male DisplayPort short cable,
OR
USB3 C male to male, long cable, then USB3 C female to female short cable, then USB3 C male to DisplayPort male short cable,

that way I can rejig the USB3 C cable for future monitors by just changing the end cables.


life is so complicated! they ought to invent a dual gender socket, where all the sockets are the same, but maybe that is impossible.

also could you recommend any SATA extender cables with clip attachments, as I cannot find any at the URL you gave in the earlier topic:

https://eustore.cablemod.com/product-category/cables/sata-data-cables/

none say they are extenders, ie male to female, maybe I havent understood the website.

they have a category of extenders, but I cannot discern SATA extenders:

https://eustore.cablemod.com/product-category/cable-extensions/


on p8 of the manual with the memory sockets, arranged A1, A2, B1, B2,
they say to use A2 first, then B2, is there any reason for such a tortuous scheme other than to create stress for their customers?

couldnt they just rejig the wires where it is left to right? eg use the one nearest the CPU first,

eg with my fridge, I put milk cartons with the earliest date outermost, where I just reach for the first available one, rather than some crazy scheme where you have to use the 2nd one then the 4th, etc!

what we call ergonomics.

I am waiting for the 1/4" torque screwdriver, which probably will arrive Monday, and should start installing the MDF table meanwhile. Will try to get a refund on the 4mm torque screwdriver.

 

[Aeacus]

button 1 is the "clear CMOS" button, this wont work then if you brick up the BIOS?

it says "Power off your computer. Press and hold the Clear CMOS button for about 5-10 seconds to reset BIOS to default values".

I dont know about today's mobos, but in my era, mobos had a watch battery on them, is this entirely for the clock, or is it for storing the "early startup" settings, eg boot drive order?

the BIOS updates, these are recorded to special memory that persists even when powered down like with a flash drive?

ie if the watch battery runs out, the BIOS isnt affected?

CMOS battery used to be used only to keep the PC's internal clock correct. But today's PCs can take the correct time from internet.

But CMOS battery also helps to store user made changes to the BIOS (e.g RAM XMP/EXPO, custom fan curve, boot priority, CPU OC etc). So, when you either clear CMOS or remove the CMOS battery for ~5mins, all user made changes to the BIOS settings will be reset to the factory defaults. But this doesn't fix the corrupt BIOS.

If the CMOS battery is empty, none of the changes you do in BIOS can be saved, so, you'll be running default factory settings at all times.

now I am a bit confused, I thought the graphics would need the highest speed, but this appears to suggest we need socket 10 for graphics, which presumably is half the speed of the other 2?

DP 1.4 maximum theoretical bandwidth is 32,40 Gbps. Since DP uses 4 lanes inside the cable, that would mean one lane outputting max 8,1 Gbps, making 10 Gbps USB 3 Gen 2 enough to carry the signal.

also who or what is ASM3241? it looks like a password!

It is dedicated host controller chip, which bridges PCI-E to USB 3.2.
Further reading: https://www.asmedia.com.tw/product/393yQ5esP7qw6NID/f32YQ14SmApn1wNA

you said in the other topic that buses are from either the chipset or the CPU, but here we have apparently a 3rd entity, the ASM3241.

If you follow the pipeline, it will still end up in the MoBo chipset.

X670 chipset diagram, also showing what CPU provides;

Source + further reading: https://www.hwcooling.net/en/amd-am5-platform-b650-x670-x670e-chipsets-and-how-they-differ/

so should I be using the 3 + 11 + 10 block before the chipset ones, or it makes no difference?

Little, if any difference. Minute difference is that chipset provided USB ports might be a tiny bit slower, compared to those that are hardwired directly to the CPU.

could you recommend a long USB 3 C cable which will handle the max video speeds of this mobo?

How long is long?

5 meters, USB type-C to USB type-C: https://www.amazon.co.uk/Marker-Thunderbolt-MacBook-Samsung-Monitor/dp/B0BCWKYNHM

To extend the cable, buy two of those and connect them together with USB type-C female to female adapter,
e.g this thing: https://www.amazon.co.uk/Maxhood-Ad...underbolt-USB-C-Female-Coupler/dp/B0BRCX33LR/

also any recommendation for a short cable to convert the other end of the USB3 C cable to DisplayPort for my monitor?

How short is short?

2 meters, USB type-C to DP 1.4: https://www.amazon.co.uk/UGREEN-DisplayPort-Thunderbolt-Compatible-MacBook/dp/B0C4DB8MLL/

as that would be female USB3 C, I dont know if I would need a gender converter USB3 cable as well?

or if I need a long male to female USB 3 cable?

because I have to go from female USB3 C socket to female DisplayPort socket,

where I thus need either:

USB3 C male to female, long cable, then male USB3 C to male DisplayPort short cable,
OR
USB3 C male to male, long cable, then female USB3 C to male DisplayPort short cable,
OR
USB3 C male to male, long cable, then USB3 C female to female short cable, then USB3 C male to DisplayPort male short cable,

that way I can rejig the USB3 C cable for future monitors by just changing the end cables.

Extenders are male to female.
Direct connection cables are male to male.

What i linked above is 3rd option you wrote.

It is really up to you which cables you buy and which connection method you use between the cables.

life is so complicated! they ought to invent a dual gender socket, where all the sockets are the same, but maybe that is impossible.

Not impossible.

The current male and female connectors are present because there needs to be some way to hold the two ends together. And that the male and female connectors are doing, where two ends are held together on friction basis.

Now, for two ends to have same connector, there must be some other method to hold the two ends together, other than the friction (male and female). And there is: magnetic connectors are one such connector, already used with some PC case fans. Also, smart watches use magnetic connection to connect to the data/charging cable. Maybe some more, but these two came right into my mind.

also could you recommend any SATA extender cables with clip attachments

Could not find SATA male to female extender that has metal clip on the male connector.

on p8 of the manual with the memory sockets, arranged A1, A2, B1, B2,
they say to use A2 first, then B2, is there any reason for such a tortuous scheme other than to create stress for their customers?

There is no hard rule how to wire RAM channels to the RAM slots, but most commonly used is (counting from CPU) A1, A2, B1, B2, whereby when using single DIMM, slot it into 2nd slot from CPU (A2). When you have two DIMMs, then into 2nd and 4th slot from CPU (A2, B2).

Here's more in-depth about how CPU is connected to RAM slots;

View: https://www.youtube.com/watch?v=3vQwGGbW1AE

 

[35below0]

The CMOS battery generally won't run out, but if it does the effect is the same as removing it deliberately and replacing after ~30s. It will clear CMOS.
Another way to clear CMOS is to short two pins on the motherboard by touching them with a screwdriver for 5-10s.
Having a button to press is far more convenient.

And yes, BIOS settings are saved to the motherboard itself and persistant. Clearing CMOS reverts BIOS settings to their default configuration. Every version of BIOS has it's default configuration. After updating BIOS it is advisable to also reset to defaults. This can be done in BIOS itself. Clearing CMOS is done if BIOS cannot be used. Or seen, as sometimes the computer works just fine but display is unavailable.

Bus lanes are broadly split into two groups, the CPU side and the chipset side. Nothing more.
There is only so much you can stuff onto the motherboard before different devices are forced to share lanes/controller. This affects GPUs and NVMe drives, but it's not as complicated as it may seem. Pretty much the GPU goes into the fastest PCIe slot and that's that, unless you have two or three GPUs.
With NVMe drives, you have a choice between a CPU side M.2 slot or the chipset side M.2 slot(s). And again, you pretty much insert the drive into the CPU slot and call it a day.

The memory slots are not intuitive, i'll give you that. But because motherboards are designed by engineers, they work and work well. If common sense were allowed to have a say, motherboards would be far more sensible and also far more useless. Possibly obsolete.
Follow the manual and plug RAM into A2 and B2. Don't sweat the small stuff.

 

[Richard1234]

before the reply, I have completed constructing the MDF table for constructing the PC!

http://www.directemails.info/tom/table.jpg

the Travis Perkins shop wouldnt give any advice, but a customer did! he selected out some 3.5mm x 16mm screws which would suffice for the 18mm MDF, explaining that some of the length is used up at the table leg plate.

and those screws were supplied with the right drill bit in the box. he said also to drill a pilot hole through the MDF and gave me a free one!

I'll discard those screws when I uninstall to prevent them deteriorating, as the box contains a huge amount.

I set then ungauged torque of the drill to minimum which seemed to work well.

I then shunted all the PC components to the table, and noticed the MSI dragon emblem on the Geforce RTX 4060, which has Nvidia on it. are MSI and Nvidia and Geforce the same firm?

CMOS battery used to be used only to keep the PC's internal clock correct. But today's PCs can take the correct time from internet.

taking time from the internet is very unsatisfactory as I usually boot up without the internet, I only connect to the internet if I am using the internet! also what if the internet is down, or the modem is kaput?

one reason for having the correct time is for files to have the correct timestamps. eg with my many different projects, I can tell when I was working on some stuff from the timestamps. ALSO timestamps enable me to locate files where I have forgotten the filename or the directory, eg if I was working on a file a few weeks ago, I can just navigate for the 2024 timestamps.

But CMOS battery also helps to store user made changes to the BIOS (e.g RAM XMP/EXPO, custom fan curve, boot priority, CPU OC etc). So, when you either clear CMOS or remove the CMOS battery for ~5mins, all user made changes to the BIOS settings will be reset to the factory defaults. But this doesn't fix the corrupt BIOS.

so if you change the CMOS battery really quickly the settings will all be preserved?

only problem is the battery runs out first, then the timestamps cause problems, and then you change the battery!

If the CMOS battery is empty, none of the changes you do in BIOS can be saved, so, you'll be running default factory settings at all times.


DP 1.4 maximum theoretical bandwidth is 32,40 Gbps. Since DP uses 4 lanes inside the cable, that would mean one lane outputting max 8,1 Gbps, making 10 Gbps USB 3 Gen 2 enough to carry the signal.

ok, good explanation!

It is dedicated host controller chip, which bridges PCI-E to USB 3.2.
Further reading: https://www.asmedia.com.tw/product/393yQ5esP7qw6NID/f32YQ14SmApn1wNA


If you follow the pipeline, it will still end up in the MoBo chipset.

X670 chipset diagram, also showing what CPU provides;

very interesting diagram!

the diagram says the ASM4242, whereas the mobo says ASM3241, is the above diagram for a different related mobo?

when you say host controller, what is regarded as the "host", in colloquial english a "host" or "hostess" has "visitors" who they look after, the "host" here is whom? and the "visitors" are whom? the USB sockets? the peripheral hardware?

Source + further reading: https://www.hwcooling.net/en/amd-am5-platform-b650-x670-x670e-chipsets-and-how-they-differ/



Little, if any difference. Minute difference is that chipset provided USB ports might be a tiny bit slower, compared to those that are hardwired directly to the CPU.


How long is long?

5 meters, USB type-C to USB type-C: https://www.amazon.co.uk/Marker-Thunderbolt-MacBook-Samsung-Monitor/dp/B0BCWKYNHM

5m is plenty, 4m would also be alright

To extend the cable, buy two of those and connect them together with USB type-C female to female adapter,
e.g this thing: https://www.amazon.co.uk/Maxhood-Ad...underbolt-USB-C-Female-Coupler/dp/B0BRCX33LR/

the problem with that kind of adapter, is you have 3 items connected at one place, so there is more risk of one getting dislodged via twisting at the adapter, it is better to have a short female to female cable, that way you separate that into 2 connections of 2 items.

I had this problem when I attached a USB 3 C to A adapter without cable to my 2010 mobo, where it started to detach at one end,

it is better to convert via a short cable than via a solid converter

How short is short?

2 meters, USB type-C to DP 1.4: https://www.amazon.co.uk/UGREEN-DisplayPort-Thunderbolt-Compatible-MacBook/dp/B0C4DB8MLL/

2m is a bit too long to be short! I think 1m would be enough

would the electric signals sustain for 5m + 2m? or is it optical?

Extenders are male to female.
Direct connection cables are male to male.

ok, wasnt aware of that

What i linked above is 3rd option you wrote.

It is really up to you which cables you buy and which connection method you use between the cables.


Not impossible.

The current male and female connectors are present because there needs to be some way to hold the two ends together. And that the male and female connectors are doing, where two ends are held together on friction basis.

Now, for two ends to have same connector, there must be some other method to hold the two ends together, other than the friction (male and female). And there is: magnetic connectors are one such connector, already used with some PC case fans. Also, smart watches use magnetic connection to connect to the data/charging cable. Maybe some more, but these two came right into my mind.

magnetic is better than friction, because friction based things eventually wear out, maybe one day hardware will communicate wirelessly, where no sockets at all, like with bluetooth, except the bluetooth dongles.

eg a friend of mine has a TV box which connects via the modem via 2 TPLink gizmos, but my one connects wirelessly. Possibly theirs does too, I might have a study of that.

as long as the magnetism also doesnt wear out! potentially it could be remagnetised.

I have an exercise bike with magnetic resistance, much better than frictional resistance which wears out.

Could not find SATA male to female extender that has metal clip on the male connector.

could you find any which are female to female?

as I could connect:

clip-male ---- clip male female---female

where the composite is clip-male ---------- female

you said normal cables are male to male, so presumably the normal clip ones are then clip-male to clip-male?

as explained earlier, a female to female cable would be preferable to a solid female to female adapter.

There is no hard rule how to wire RAM channels to the RAM slots, but most commonly used is (counting from CPU) A1, A2, B1, B2, whereby when using single DIMM, slot it into 2nd slot from CPU (A2). When you have two DIMMs, then into 2nd and 4th slot from CPU (A2, B2).

so its just to conform with other mobos?

I suppose 2nd then 4th slot means you have some distance between the 2 which would be easier to install,

eg using up all my sata sockets on my 2010 mobo, it is tricky to detach the inner ones, the outermost ones are ok, the others are a right nuisance.

but A1 then B2 would be even bigger distance!

I think the entire ATX specification needs modernising, maybe to a BTX format!

where they can rethink the spacings to be appropriate for modern hardware, with bigger distances where necessary, rather than wasting bandwidth on space filler sockets

ATX must be from decades ago.

with all things, after some years you need to rethink the arrangement,

I think the mobo should supply the PCI back panel, rather than the tower, that way they can do bespoke arrangements.

 

[35below0]

The computer would have to be offline from the internet for a very long time before the clock went off. If it even happened. I'm not sure which service the OS gets the time from.

Anyway, you can set the clock in BIOS manually and that will probably stay correct for the PC's life.


Geforce is the brand name. It's been used for a long time now. Nvidia designs the GeForce series though it doesn't manufacture the cards. That is done by parties such as MSI, Gigabyte, AsRock, Gainward, and others. These companies also manufacture and market AMD GPUs. They are hardware manufacturers and make a range of components including graphics cards, power supplies, motherboards, etc. What they don't do is design and create actual graphic cards (Intel is the exception). Nvidia and AMD do that as well as supply the drivers, and they let others build the things according to Nvidia/AMD specs.
Cards made by different manufacturers are also slightly tuned from the base design. Some are slightly overclocked, the cooling solution may be different, and so on.

The RTX 4060 is designed by Nvidia who also owns the IP. MSI manufactured, marketed and sold the particular model you own.


The battery will not run out for a very very long time. They easily last 10+ years.
If you remove the batter yourself, and then quickly replace it, it might not clear CMOS so you will not revert to default BIOS configuration. This is what you normally want to do when you remove the battery.
Do note that removing the battery is a pain in the nether regions and it is easy to damage parts of the motherboard if it's done carelessly. It is much easier to short pins with a screwdriver but even that can be risky if you accidentally touch other stuff.
Having a dedicated clear CMOS button to press is much more convenient.

It's important to remember what Aeacus said above. If a BIOS update goes awry and the motherboard is stuck without a functioning BIOS, clearing CMOS will have no effect. It's only effect is to set BIOS to default configuration, but the BIOS has to be there in the first place and be functional. A "bricked" motherboard cannot be restored without flashing/reinstalling BIOS itself, but that may not even be possible!
Your motherboard has dual BIOS so you'll be safe in case BIOS is corrupted as you will have a "spare" so to speak.


As for the BTX format, the less said of that disaster, the better.
ATX may be long in the tooth but it is extremly adaptible and extremly compatible. You introduce multiple standards of standards and specifications, and the poor consumer will be more confused than ever. And manufacturers will see their profit margins shrink as they spread themselves thin trying to cater to more and more designs, each one more niche than the one before.
As it is, it's pretty complicated already. There's 3-4 motherboard standards, times two for the two CPU manufacturers, possibly even more complicated in case of an AMD socket and it's relation to DDR4 and DDR5 ram.
Intel also likes to complicate things further by changing socket designs relatively frequently. While it doesn't affect the ATX layout much, it does make older sockets obsolete quickly and with them older CPUs.
Then there's OEMs like HP and Dell who will insist on using proprietary in-house designs for motherboards, power supplies and case layouts.
It's a jungle. Adding a revolution to the mix is not going to improve things.

A revision of the ATX design with enough lead time to allow manufacturers and consumers to "move on" to the new spec is the best way to do it. Small steps. Plenty of time to catch mistakes and screwups.

Occassionally someone may come up with a niche product or a clever idea, but not the mainstream players. If the idea is good, it will move into the mainstream though. And if the idea is bad, or if it is trying to revolutionize things that don't need revolutionizing, then it will fare as well as the BTX did. Badly.

 

[USAFRet]

I think the entire ATX specification needs modernising, maybe to a BTX format!

Why should we regress to a much worse format?

 

[Aeacus]

I then shunted all the PC components to the table, and noticed the MSI dragon emblem on the Geforce RTX 4060, which has Nvidia on it. are MSI and Nvidia and Geforce the same firm?

Not the same firm. Instead AIB (Add-In Board partner).
Every company that isn't Nvidia or AMD but sells their own made GPUs, is AIB. That includes, but isn't limited to: MSI, Asus, Gigabyte, Zotac, Palit, AsRock, Colorful and more.

Nvidia makes the GPU chip + specifies the components on the PCB (e.g how much VRAM). AIB (e.g MSI) then buys the chip and rights to make a GPU. Now, AIB can customize their own PCB, but still keeping in the specs of defined by Nvidia. Since Nvidia is quite strict with their specifications, not much can be improved. Two main things that AIB GPUs do better than Nvidia FE (Founders Edition) GPUs do: are power delivery and cooling. So, AIB GPU (aftermarket GPU) can be OCd more easily, to higher clock levels and it also has better cooling on it (lower temps, less noise). As of how GPU looks (eyecandy) that too is up to AIB to decide. But since GPU uses Nvidia chip, it has to include Nvidia name on the GPU or on the box.

Btw, GeForce is product line name from Nvidia, indicating consumer GPUs.
On AMD side, Radeon is product line name for consumer GPUs.

I only connect to the internet if I am using the internet! also what if the internet is down, or the modem is kaput?

I'm not entirely sure, but Windows 11 may not work at all if the PC isn't connected to the internet. Heck, you may not be even able to install Win11 without being connected to internet.
Win10 also throws a fit when it isn't connected to internet but is more lenient about it.

so if you change the CMOS battery really quickly the settings will all be preserved?

Yes, since it take some time to clear CMOS once you remove the battery.

the diagram says the ASM4242, whereas the mobo says ASM3241, is the above diagram for a different related mobo?

Well, above diagram is X670 chipset. You have X670E chipset. I did try to find X670E chipset diagram but wasn't able to. Also, the chip itself can vary a bit depending on MoBo manufacturer.

when you say host controller, what is regarded as the "host", in colloquial english a "host" or "hostess" has "visitors" who they look after, the "host" here is whom? and the "visitors" are whom? the USB sockets? the peripheral hardware?

A host controller is a separate programmable microcontroller.
Whereby microcontroller is a compact integrated circuit designed to govern a specific operation in an embedded system. A typical microcontroller includes a processor, memory and input/output (I/O) peripherals on a single chip.

would the electric signals sustain for 5m + 2m? or is it optical?

If it is under 10 meters, it's good. Passive video signal cable can do ~15 meters (e.g HDMI), any longer that that would need active signal boost somewhere in-between.

maybe one day hardware will communicate wirelessly, where no sockets at all, like with bluetooth, except the bluetooth dongles.

Signal mixing is one issue that i can think of.

Another is the fact that electricity can't be transferred wirelessly. Current tech, at best, is when device (e.g smart phone) sits on the dedicated charge pad, to be charged "wirelessly". But not on the distances of regular wi-fi, which would be needed if you'd want your monitor to wirelessly connect to your PC to show image.

could you find any which are female to female?

This thing is rare as well. Did find one, but not on amazon,
link: https://www.octo24.com/en/InLine-27700A-InLine-SATA-Cable-Extension-Adapter::486522.html

Though, did find SATA male to female with clip, but it will extend both, data and power,
amazon UK: https://www.amazon.co.uk/Xiatiaosann-22-Pin-Extension-Sleeved-Optical/dp/B0CD7QNYN5/

with all things, after some years you need to rethink the arrangement,

If it ain't broke - don't fix it.

Current ATX standard works just fine. I see no reason why to reinvent the wheel.

I suppose 2nd then 4th slot means you have some distance between the 2 which would be easier to install,

but A1 then B2 would be even bigger distance!

You don't install/remove RAM on daily basis for the best slots to be in order when counting from CPU. Most of the times, you install the RAM once and that's it. Maybe in the future, you'll add 3rd and 4th DIMM, or replace the two you have with bigger capacity ones. Still, not enough reason to retrace memory lanes for the minute convenience you'll get out from it.

Nvidia designs the GeForce series though it doesn't manufacture the cards.

Doesn't manufacture GPUs? Ever heard of FE (Founders Edition) versions?

Nvidia, besides selling the GPU chips to AIBs, is keeping the best binned chips for themselves, to put into FE versions and to compete with AIBs. Nvidia is quite arrogant in that regard actually. Prime example would be EVGA and why EVGA, who used to be longest AIB partner for Nvidia, completely stopped being Nvidia partner starting from RTX 40-series.

 

[35below0]

I'm not entirely sure, but Windows 11 may not work at all if the PC isn't connected to the internet. Heck, you may not be even able to install Win11 without being connected to internet.
Win10 also throws a fit when it isn't connected to internet but is more lenient about it.

It is possible to install Win 11 without an internet connection. Some users do this deliberately, to avoid creating a Microsoft account.

Win 11 complains about no internet connection if it cannot find one. It is designed to install with a working internet connection and will force the user to create an account if it can. But it doesn't require a connection as such.
This personalization stuff can later be removed, except for the account itself.

It is almost exactly like setting up a new mobile phone works. The user is forced to create an Android/Apple account, although how much the user has to care about that is up to them. A lot of people ignore the account or have no idea what the password is. Until they need to migrate or access their data at least.

As for Win 11, it is designed to install and work with an internet connection. I see no reason to recommend against this. After setting up Windows, the OS will use the internet to gather relevant drivers, and to download and install OS updates. A restart or two may be neccessary.

Once the setup business is finished, it's ok to disconnect and connect at will. The OS does not require a connection in order to function.

Doesn't manufacture GPUs? Ever heard of FE (Founders Edition) versions?

Nvidia, besides selling the GPU chips to AIBs, is keeping the best binned chips for themselves, to put into FE versions and to compete with AIBs. Nvidia is quite arrogant in that regard actually. Prime example would be EVGA and why EVGA, who used to be longest AIB partner for Nvidia, completely stopped being Nvidia partner starting from RTX 40-series.

My answer was very general. Intel makes it's own cards, Nvidia and AMD do not. Yes there's Founders editions but the overwhelming majority of Nvidia cards out there are not made by Nvidia itself. I'm not sure if Intel is partnering with anyone but it's possible there is interest.

 

[Aeacus]

As for Win 11, it is designed to install and work with an internet connection. I see no reason to recommend against this. After setting up Windows, the OS will use the internet to gather relevant drivers, and to download and install OS updates. A restart or two may be neccessary.

Speaking of it, decided to throw out the Win11 install guide made by our very own Darkbreeze, with pics and all,
link: https://forums.tomshardware.com/threads/windows-11-clean-install-tutorial.3831442/

It's for OP and for anyone else who happens to read this topic.

Also, Win10 installation guide by Darkbreeze as well, again with pics and all,
link: https://forums.tomshardware.com/faq/windows-10-clean-install-tutorial.3170366/

So, @Richard1234 , when it comes time to install OS in your build, you can then read the very well made in-depth guide on how to install the OS. This way, i don't have to retype it all.

I'm not sure if Intel is partnering with anyone but it's possible there is interest.

Would love to see Intel partnering with EVGA since EVGA is one of the oldest GPU manufacturers and would help Intel immensely in making their GPUs.

There were talks of public after EVGA ended their partnership with Nvidia, that EVGA should partner with Intel but EVGA didn't disclose anything about it to the public. Maybe, behind the scenes, EVGA is helping out Intel. We don't know.

Though, would love to see EVGA and K|NGP|N back in action. K|NGP|N helped to produce many innovative GPUs and MoBos for EVGA.

 

[Richard1234]

Why should we regress to a much worse format?

arg! didnt realise the label was already used!

what I meant was a brand new industry format based on discussion by the different participators.

in Germany, I noticed that when a work is done, they have a big discussion involving everyone, customer, workers, the boss, everyone. then they arrive at a consensus, and this is the work done.

whereas in Britain, they will have a 5 minute discussion with the customer and begin the work and make bad decisions.


I think there is a general problem with hardware engineers that they are sealed off from everyone else!

 

[Richard1234]

The computer would have to be offline from the internet for a very long time before the clock went off. If it even happened. I'm not sure which service the OS gets the time from.

Anyway, you can set the clock in BIOS manually and that will probably stay correct for the PC's life.

what I meant was if there is no mobo clock and the time is taken from the internet.

As for the BTX format, the less said of that disaster, the better.

I just invented the name BTX! I didnt expect the name to be in use! so I wasnt referring to this pre-existing format!

I just assumed that as the standard is ATX, that BTX wouldnt be in use! I'd expect the current format to be the last letter used!

ATX may be long in the tooth but it is extremly adaptible and extremly compatible. You introduce multiple standards of standards and specifications, and the poor consumer will be more confused than ever. And manufacturers will see their profit margins shrink as they spread themselves thin trying to cater to more and more designs, each one more niche than the one before.
As it is, it's pretty complicated already. There's 3-4 motherboard standards, times two for the two CPU manufacturers, possibly even more complicated in case of an AMD socket and it's relation to DDR4 and DDR5 ram.
Intel also likes to complicate things further by changing socket designs relatively frequently. While it doesn't affect the ATX layout much, it does make older sockets obsolete quickly and with them older CPUs.
Then there's OEMs like HP and Dell who will insist on using proprietary in-house designs for motherboards, power supplies and case layouts.
It's a jungle. Adding a revolution to the mix is not going to improve things.

what I meant is with proper discussion, an agreed cross industry standard, and not just one firm out there pushing a new idea. but a proper discussion by everyone as regards the tower case and mobo side. maybe via a forum created to discuss the matter.

Spoiler: unexpanded to prevent proliferation of the thread as a bit off topic

not discussing the memory socket formats, but just the geometry of the external sockets relative to the case.

but you could also have say which socket to use first,



currently there seem to be slots at fixed locations on the tower, which the mobo has to align with, but they then have to create PCI sockets which will never be used just to give the graphics card enough space.

this is a waste of bandwidth

instead it would be much better to make those slot locations part of the mobo and not the tower case, where the mobo supplies its own panel for that zone. You could still keep a detachable panel on the tower for ATX etc. The way mobos have back panels for various sockets, just asking eg to do this for PCI sockets also. its not rocket science, you could literally file off the tower ones.

the mobo can then put the socket wherever is expedient in a zone, and the next socket wherever is expedient, and supply a fixed size overall panel for that zone.

things like memory socket formats is a different matter, and again is best done with cross industry agreement.

its a problem with electric vehicles, that Tesla does one thing, other firms another, and there might just be Tesla chargers in one place. Much better for the entire industry to come up with one standard which everyone uses. electricity is ultimately a voltage and a frequency, must be possible to have one format for all cars maybe with different options, the same way you have petrol as the main option for cars, and diesel for bigger vehicles.

like with batteries, we have AA and AAA, C etc. Sony got mired with their Betamax format, where everyone else went for VHS. after that Sony always ensures its standards are open, eg CD, DVD done in conjunction with Philips.

HP and Dell doing their own incompatible things, is exactly what I am opposing, and is also the problem with Tesla and Apple, where they deliberately shut out others.

at school, we were taught the social and economic history of Britain, and one topic was the importance of standardisation, which historically happened for screws and bolts, where they decided to standardise these, which meant any firm could manufacture these, and any firm use them. originally each firm created their own format and had to arrange manufacture of these. I think they realised this in the 1800s! this is where I am disappointed by Tesla, HP, Dell, Apple for flouting this early lesson, leading to jack of all trades manufacturing such as Apple.

A revision of the ATX design with enough lead time to allow manufacturers and consumers to "move on" to the new spec is the best way to do it. Small steps. Plenty of time to catch mistakes and screwups.

you have to involve all manufacturers, users, programmers, etc in an open design. And not just Intel spewing out whatever they want, eg right from the start I thought USB sockets were a bad design as not symmetric, but almost symmetric so are very tricky to use the right way round. whereas say british power plugs are highly asymmetric so no confusion. either make the socket outright symmetric, or outright asymmetric. but the USB A socket is neither here nor there, every time I insert one it is stress determining of its the right way round.

this kind of long running design gaffe wouldnt happen if you had open discussion, users would immediately shoot down this stupid design aspect.

in Germany the power sockets are fully symmetric, this concept of symmetric sockets is well ancient, Intel have no excuse!

Occassionally someone may come up with a niche product or a clever idea, but not the mainstream players. If the idea is good, it will move into the mainstream though. And if the idea is bad, or if it is trying to revolutionize things that don't need revolutionizing, then it will fare as well as the BTX did. Badly.

I only used BTX by coincidence, I was trying to allocate an unused name! looking online, BTX is Intel, so its not a surprise it was badly thought out.

 

[Richard1234]

the 1/4" torque screwdriver arrived whilst doing these replies! but have run out of time today, so I will try to work on installing on Tuesday. Tuesday also they are collecting the 4mm screwdriver for a refund, where the firm arranged UPS to collect this.

A host controller is a separate programmable microcontroller.
Whereby microcontroller is a compact integrated circuit designed to govern a specific operation in an embedded system. A typical microcontroller includes a processor, memory and input/output (I/O) peripherals on a single chip.

this is slightly over my head! maybe some specific example hardware configs to illustrate the idea?

If it is under 10 meters, it's good. Passive video signal cable can do ~15 meters (e.g HDMI), any longer that that would need active signal boost somewhere in-between.

ok, this is useful info.

also where I said the max video rate the mobo can do, I really should have said the max video rate either the mobo or RTX 4060 can do!

Signal mixing is one issue that i can think of.

you could have a selector on devices for different fixed frequencies. OR transmit different selectable frequencies over the same higher carrier frequency. where different devices filter out their frequency. unless that would require a problematic carrier frequency.

with modems, Windows gives me a list of local modems, so the hardware has the means to select out one from many. not sure how it does that?

Another is the fact that electricity can't be transferred wirelessly. Current tech, at best, is when device (e.g smart phone) sits on the dedicated charge pad, to be charged "wirelessly". But not on the distances of regular wi-fi, which would be needed if you'd want your monitor to wirelessly connect to your PC to show image.

the power would need to be local, eg my handshake mouse has an inbuilt battery charged via a USB C cable.

wireless would just be to transmit information, not power! maybe Nicola Tesla had ways to transmit power wirelessly!

This thing is rare as well. Did find one, but not on amazon,
link: https://www.octo24.com/en/InLine-27700A-InLine-SATA-Cable-Extension-Adapter::486522.html

the problem with this one is you have 3 gizmos connected at one point, where there is a risk of twisting, but maybe the clip will prevent twisting. for USB A to C solid converter, I found twisting to be a problem.

maybe I will try these to see if they are ok.

Though, did find SATA male to female with clip, but it will extend both, data and power,
amazon UK: https://www.amazon.co.uk/Xiatiaosann-22-Pin-Extension-Sleeved-Optical/dp/B0CD7QNYN5/

could be overspecified!

If it ain't broke - don't fix it.

Current ATX standard works just fine. I see no reason why to reinvent the wheel.

it is broke, in that you need bogus PCI sockets because not enough space for the graphics card!

instead much better just to have that PCI socket panel on the tower be now supplied by the mobo, where it can be bespoke, like with the back panel for USB and other sockets.

very simple enhancement, and towers can supply a removable panel to be compatible with old mobos.

we mustnt just blindly accept whatever things are currently. but it is best to change things in a cooperative way.

 

[Richard1234]

one further question:

what is the max total length for SATA cables?

 

[helper800]

one further question:

what is the max total length for SATA cables?

One meter.

 

[Aeacus]

in Germany, I noticed that when a work is done, they have a big discussion involving everyone, customer, workers, the boss, everyone. then they arrive at a consensus, and this is the work done.

Can't do that with MoBo (or any electronics that is sold to consumers in that matter), since there is no way you can get entire world behind one table, to discuss how the new method/design would appeal to everyone.

I think there is a general problem with hardware engineers that they are sealed off from everyone else!

Engineer may make a good product, but it comes down to company (e.g CEO/shareholders) that decide if the product is to be produced or not.

what I meant is with proper discussion, an agreed cross industry standard, and not just one firm out there pushing a new idea. but a proper discussion by everyone as regards the tower case and mobo side. maybe via a forum created to discuss the matter.

In PC hardware world, there are some big players who like to do their own thing. Few examples:

Nvidia - +12VHPWR connector and it's revision: 12V-2x6.
This is only Nvidia thing and found on RTX 40-series. It started with RTX 30-series. Whereby i think Nvidia made this connector to save space. Before that, and what AMD still uses with their Radeon GPUs + Intel with their ARC GPUs, are the industry standard PCI-E power connectors. Either 6-pin or 8-pin. But since latest and beefiest GPUs would require 4x 8-pin connectors on GPU, it would take up a lot of precious real estate on GPU PCB. So, i think this is the reason why Nvidia did their own thing and released +12VHPWR connector, which has serious issues (connector melting).

Intel - ATX12VO.
Currently PSU takes 120V/240V AC in and outputs +12V DC, +5V DC and +3.3V DC to MoBo and other components. Now, Intel came out with an idea, where PSU only outputs +12V DC and it's up to the MoBo to convert the +12V DC into +5V DC and +3.3V DC. Essentially moving the DC-DC conversion (voltage regulation) from inside the PSU to ontop of a MoBo.

I find it bad idea on so many levels.
* For one, real estate on MoBo is already at premium, making me ask, to where the DC-DC converter is supposed to go.
* Then, of course, is the matter of efficiency, ripple and noise. Currently PSUs have to meet strict standards, so that they are safe to use. PSU manufacturers have years of experience in making PSUs. But this would move the responsibility to MoBo manufacturer and MoBo manufacturer does not have the knowledge to make good PSUs. E.g i do not know any MoBo manufacturer, who, at current date, also produces their own PSU and that PSU is either good or great quality. MSI, EVGA, Gigabyte, NZXT and other MoBo manufacturers, who do sell PSUs under their brand name, do not make the PSU themselves. Instead, they outsource the PSU manufacture to PSU OEMs (e.g Seasonic, Super Flower, Great Wall, CWT, HEC/Compucase).
* Another thing is, that in this case, PSU would get both smaller and cheaper, since it only needs to output +12V DC. But in turn, MoBo gets bigger (it has to house DC-DC converter) and more expensive as well. Currently, price wise, high-end PSU costs about the same as mid-tier MoBo. But in terms of ATX12VO, PSU could cost peanuts while MoBo costs a fortune.
* Cable management will go worse as well, since for e.g SATA power (or MOLEX), which has +5V DC and currently comes from PSU, will have connector on the MoBo itself.

At current date, ATX12VO is used with some prebuilt PCs, which already houses proprietary MoBo and PSU. Now, Intel's idea was to streamline power delivery with prebuilt proprietary PCs, but there's no telling that this won't flow over to the current DIY segment of PCs. Heck, you can already buy MoBos that accept only +12V DC.

Further info about ATX12VO if interested:

View: https://www.youtube.com/watch?v=Zc9oRKexV_s

this is slightly over my head! maybe some specific example hardware configs to illustrate the idea?

This very chip we talk about, enables MoBo to use USB and PCI-E protocols from the same output port. Normally, USB and PCI-E are separate, with their own separate "data flow" to the system. Now, this specific chip acts as a bridge between USB and PCI-E.

In layman's terms, it enables data from USB port to travel to PCI-E protocol, which is much faster, compared to the USB, where USB data packages have to wait in line.

For analogy example;
Let's say there are two parallel roads going in same direction into the city (into the PC), without ever intersecting with each other.
One road is highway, named PCI-E and has speed limit of 90 KM/h.
Another road is street, named USB and has speed limit of 60 KM/h.

Now, data can travel on both roads, but depending from where they came from, they are either forced to travel on the PCI-E road or on the USB road. Flow on USB road is slower and there can be traffic jams as well due to rush hour.

But if you could build an on-ramp from USB road to PCI-E road, then it alleviates the traffic load on USB road and also would make the traveling data faster, since they can now travel on faster PCI-E road, to the city (into the PC).

ASM3241 chip is that on-ramp between USB and PCI-E, connecting both.

Did you get it now?

I really should have said the max video rate either the mobo or RTX 4060 can do!

Here, i need to know the specific GPU you're getting, because i need to look up the ports on GPU. But GPUs can do easy 8K resolution, far more than what MoBo is capable of outputting.

you could have a selector on devices for different fixed frequencies. OR transmit different selectable frequencies over the same higher carrier frequency. where different devices filter out their frequency. unless that would require a problematic carrier frequency.

There is finite amount of different frequencies a device can use and it is already over saturated.

E.g i can tap into my neighbor's temperature reading unit, since i think we both bought the same Sencor weather station. And i'm sure my neighbor can do the same, reading my weather probes from their device.
This, of course, brings another issue, if ALL devices are wireless, how can you protect the data when it is carried over air? E.g you are watching something from your wireless monitor, while neighbor also has wireless monitor and can tap in, and also watch the very same thing as you're watching, without you ever realizing it.

with modems, Windows gives me a list of local modems, so the hardware has the means to select out one from many. not sure how it does that?

Due to different SSID.

wireless would just be to transmit information, not power! maybe Nicola Tesla had ways to transmit power wirelessly!

So, what you're asking is wireless peripherals, that work over wi-fi or bluetooth? E.g wireless mice, wireless KB, wireless headset. But not only those, but wireless monitor as well?

If so, then monitor takes on the order of up to 100W and would quickly drain any reasonably sized battery. So long as you have to run a power cable to the monitor, there's not much benefit to be had by sending the video over a wireless connection.

instead much better just to have that PCI socket panel on the tower be now supplied by the mobo, where it can be bespoke, like with the back panel for USB and other sockets.

very simple enhancement, and towers can supply a removable panel to be compatible with old mobos.

There are flaws in that idea.

Different PC cases have different amounts of PCI slots. From as little as no PCI slots or two PCI slots, up to 10 or even 12.
Now, are you saying, that MoBo must include separate brackets for: 2 PCI slots, 3 PCI slots, 4 PCI slots, 5 PCI slots, 6 PCI slots, 7 PCI slots, 8 PCI slots, 9 PCI slots, 10 PCI slots, 11 PCI slots, 12 PCI slots? Just to accomodate all PC case sizes? That's 10 different brackets, where user would only use one, depending on how big their PC case is, while the rest, 9 brackets, are completely wasted.

Current design is much better, where there are individual PCI slot covers, whereby one can freely remove only those PCI slot covers they need, while keeping the rest. E.g like i had to do, whereby i removed 2nd and 3rd cover (to house GPU) and also 9th cover, to reroute one cable from inside to outside.

Spoiler: Picture proof

what is the max total length for SATA cables?

Yes, 1 meter. Min is 30cm. Most SATA data cables are around 50cm.

 

[Richard1234]

One meter.

thanks for the info! it does constrain what I am trying to do.

 

[Richard1234]

one question before the reply:

I have some hardboard, where one side is "rough", see this photo:

http://www.directemails.info/tom/hardboard.jpg

is that static resistant?

what about an ordinary MDF surface?

I dont know if MDF surfaces are a standard material, I just buy the plain ones like in the earlier photo of the table.

Can't do that with MoBo (or any electronics that is sold to consumers in that matter), since there is no way you can get entire world behind one table, to discuss how the new method/design would appeal to everyone.

I disagree 100%!

Spoiler: expand for VERY OFF TOPIC explanation why you dont need to involve everyone in the world for a good decision

you dont need to involve the entire world, there is this thing called statistics, which is that to evaluate something you only need a "sample".

so you only need representatives of each group of players, whether it be mobo manufacturers, tower manufacturers, users, programmers, etc. what are called "representative samples".


eg if you want to find the average height of 50 year old men in England, you dont need to measure every 50 year old man!

instead you say select out 1000 selected from across the country. and the average of those will be very accurate.

this is based on the fundamental theorem of probability theory, which we were taught on a compulsory introduction to probability theory course in my first year of uni maths, which says that for any probability distribution at all, the average of a sample, converges to the "normal distribution" (e^(-x^2) rescaled and shifted appropriately to have area of 1.00) as the sample size increases, the theorem gives further info and says that the normal distribution narrows around the real average. so as the sample size increases it soon selects out the true average with tiny error. where even a sample of 20 is pretty good. I have to dig out my lecture notes for the precise wording!

so even if you had a dice with strange distribution eg 1: 30%, 2: 5%, 3: 11%, 4: 41%, 5: 1%, 6: 12%, if you throw the dice 100x, and take the average of those 100 numbers eg 1 + 4 + 3 + 6 + 5 + 4 + 3 + 1 + 2 + ...., the distribution of that average will be virtually the normal distribution centred at 3.14= 1 x .3 + 2 x .05 + 3 x .11 + 4 x .41 + 5 x .01 + 6x .12, eg a particular run might have average 3.21 or something like that. it is so accurate that in business, science and medicine, averages of large samples are taken as fact.

if you ever sell the same item repeatedly on ebay, you will find the days between sales is very precise, eg when I sold photo paper, as an experiment, I sold 1 item about every 3 days, regardless whether it was HP, Epson, Canon. When I sold boxes of something like 500 comic boards, I sold about 1 a month. when I sold old era power cables, I would sell about 1 a month. Now with the power cables sometimes 2 consecutive people would buy, but only some months, and I definitely never got 4 people buy in a month. All these were experiments, not as an actual business. multiple buys are usually consecutive.

the most famous example of the theorem is pascal's triangle:
____1 (row 0)
___1 1 (row 1 )
__1 2 1 (row 2)
_1 3 3 1
1 4 6 4 1
......
where each item is the sum of the 2 above it, with the n-th row, if you divide by 2^(-n), eg 1/16 for 1,4,6,4,1 that is in fact the probability distribution for a sample of 4 spins of a coin. eg there is a probability of 4/16 for there to be 2 heads, ie for the average to be 2/4 heads.
the successive rows of Pascal's triangle, rapidly converge to the normal distribution.

in fact even a sample of just 20 will be pretty good, provided you take some steps to spread out the selection, and at the personal level even a sample of 5 is good: when I need some work doing, if I get 5 quotes, I know by the 4th quote the going price. and one step is that you shouldnt include more than 1 outlet of any firm. eg for my kitchen renovation, I got one quote from Ikea, one from B&Q, one from Everest, one from Magnet, and also quotes from some sole trader firms.

example steps for people is they arent related, they are geographically spread out, they are from different professions, etc.

so eg of you spin a coin 3 times, anything could happen, but if you do this 100 times, the number of heads wont be far from 50.

this is the basis of casinos, that the casino will profit even if the average is very slightly in their favour.

when I did my driving test at a busy city driving test centre, the tests were some strange amount of time apart, something like 43 minutes (I forget the exact amount), I asked why not say 30 minutes or 45 minutes. the instructor told me they had scrutinised the time it takes to do a test, including say traffic holding up a test, and 43 minutes maximised the number they can do in a day. which includes the time to do the administration at the end and to walk to the car.

where my test appointment was some irregular time like 2:11pm, not 2pm, not 2:15pm.

modern medicines are all based on this principle, where they do double blind placebo testing, where neither doctor nor patient knows whether their medicine is a placebo. And they use statistics to decide if a medicine is a cure. there will always be some people where a medicine will fail, but it needs to work on some

there is something called the placebo effect, where if you give 2 groups of patients a placebo, and tell the first group it is a cure, and tell the 2nd group it is a placebo, significantly more people will be cured in the first group!

this also happens if it is a genuine cure, by saying it is a placebo less people will be cured!


now you dont need statistics to cure illness or be successful at business, you can just use deduction, observation, intuition. eg I have found a lot of remedies just from observation and conjectural deduction. things like elimination diets also will locate problem foods. I have fixed problems which the doctors cannot!

the internet gives incorrect descriptions of the placebo effect saying they appear to be cured, in fact they really are cured, but it is entirely from the power of the mind. so they have to eliminate whether any curing is just by being told its a cure.

if you get really ill, recovery becomes a choice where you decide you want to recover. medical people do a lot of tricks to make you expect to be cured, which boosts success! eg at uni I talked to a guy doing medicine, and he said they were taught that the most important thing as a doctor is to reassure the patient, typically to say to the patient "oh this is quite typical for people of your age", regardless if that is true! eg "I have this pain in my right ankle", doctor: "oh this is quite normal for people of your age". the plain fact is that even the best medicines will fail on a few percent of people. this is why medical testing doesnt require 100% success, but 97% will probably do (I havent scrutinised the actual number they use, but generally with statistics 95% is regarded as success).

so for manufacturing, even if you had just 5 mobo manufacturers, 5 software firms, 20 users, just that will eliminate a lot of really stupid decisions that we see. I think Intel dont get any users at all some of their decisions are so stupid, eg the almost symmetric USB A socket. symmetric sockets have been around long before USB A eg german power sockets and bayonet light bulbs!

I have to assume also that you havent heard the idea of "betatesting" used for programming, where you dont just release some software, but you get volunteers to try out the software, which is called betatesting, and they always will give major feedback of problems. eg with Firefox, if you bookmark a URL, it gives too limited a window.

even if you just get 3 betatesters, the software will be vastly better than if you dont use betatesters.

so not only can you involve the different players, this is standard MO for software.

the big problem is hardware companies arent betatesting their designs, but modern software is majorly betatested.

now for the betatesting of hardware, it will greatly pay to carefully select the betatesters, eg if they select someone like you, the betatesting would be a lot better than if they selected me!

example of stupid design is the old hard drive jumpers, when all that could just be done with a slider switch or a rotating dial, to select different options, its not rocket science.

Engineer may make a good product, but it comes down to company (e.g CEO/shareholders) that decide if the product is to be produced or not.

Spoiler: expand for off topic comments

the way decisions are made is complicated, but ultimately you need people with the authority to make final decisions. some people wont have authority to decide, where their decisions have to be approved by a higher up.

full authority to decide can be handed down successively from the top. Be aware that not all firms have shareholders, eg I think Lidl doesnt have shareholders.

organisational structure is very complicated, and ultimately every firm is structured differently, some firms are super efficient, eg with Vision Express in the early 2000s, I walked in without appointment, had an eye test, selected my glasses and walked out with the finished product some hours later. some amazon and ebay purchases arrive the next day! others can take weeks! when I had my boiler replaced, they brought the wrong pump, they managed to source another one in literally 1 hour.

Some decisions may need shareholder approval, usually a few fund managers have a majority holding, so their approval is sufficient.

with the US, to prevent being overrun, some groups are set up who are answerable to noone. eg rogue CIA units who start dealing in guns and drugs!

In PC hardware world, there are some big players who like to do their own thing. Few examples:

Nvidia - +12VHPWR connector and it's revision: 12V-2x6.
This is only Nvidia thing and found on RTX 40-series. It started with RTX 30-series. Whereby i think Nvidia made this connector to save space. Before that, and what AMD still uses with their Radeon GPUs + Intel with their ARC GPUs, are the industry standard PCI-E power connectors. Either 6-pin or 8-pin. But since latest and beefiest GPUs would require 4x 8-pin connectors on GPU, it would take up a lot of precious real estate on GPU PCB. So, i think this is the reason why Nvidia did their own thing and released +12VHPWR connector, which has serious issues (connector melting).

this is where it hasnt been betatested properly! the Samsung smartphones that went into meltdown are a classic example of poor engineering and lack of betatesting.

Intel - ATX12VO.
Currently PSU takes 120V/240V AC in and outputs +12V DC, +5V DC and +3.3V DC to MoBo and other components. Now, Intel came out with an idea, where PSU only outputs +12V DC and it's up to the MoBo to convert the +12V DC into +5V DC and +3.3V DC. Essentially moving the DC-DC conversion (voltage regulation) from inside the PSU to ontop of a MoBo.

I find it bad idea on so many levels.
* For one, real estate on MoBo is already at premium, making me ask, to where the DC-DC converter is supposed to go.
* Then, of course, is the matter of efficiency, ripple and noise. Currently PSUs have to meet strict standards, so that they are safe to use. PSU manufacturers have years of experience in making PSUs. But this would move the responsibility to MoBo manufacturer and MoBo manufacturer does not have the knowledge to make good PSUs. E.g i do not know any MoBo manufacturer, who, at current date, also produces their own PSU and that PSU is either good or great quality.

Spoiler: expand for off topic comments

exactly, this is why standardisation is so important, where what you describe is the problem of the "jack of all trades",
where one firm tries to make everything.

best to leave power supply to the power supply specialists!

it shows also the arrogance of Intel, that they think they know better than anyone.

a really great example of what happens when you bring in other players is RISC chips, where when they started studying the output of compilers (ie programs which convert an ascii program written in a language eg Pascal or BASIC, to an executable which the OS can run), they found that compilers only used a small subset of available instruction codes, eg they didnt use "binary coded decimal". so they then junked all the stuff that nobody used, leading to much less electronics, and then regularised these remaining instructions which then became RISC, eg MIPS is the classic case.

they regularised both the format of instructions, AND the execution of instructions, where the latter led to I think a 4x speed up for MIPS at a given technology level. MIPS was 4x as fast as Motorola or Intel at the same level of technology. today's Intels use a RISC core. where the CPU is fake, and is software above an undocumented inner RISC CPU.

MSI, EVGA, Gigabyte, NZXT and other MoBo manufacturers, who do sell PSUs under their brand name, do not make the PSU themselves. Instead, they outsource the PSU manufacture to PSU OEMs (e.g Seasonic, Super Flower, Great Wall, CWT, HEC/Compucase).

Spoiler: expand for off topic comments

this is a form of "division of labour", which is to split a work into different parts, and different people work on different parts. there are 2 forms of division of labour, with the industrial revolution, the division of labour is deskilled, where each work is split into parts which anyone could do, where they can then pay much less! eg McDonalds fast food is totally deskilled. versus what you describe which is where you bring in specialists for each part, eg the company lawyer, the company statistician, the janitor, the salesmen, etc. where you have to pay more, but each part is perfect. eg the hotel restaurant, where they will have an adept chef.

* Another thing is, that in this case, PSU would get both smaller and cheaper, since it only needs to output +12V DC. But in turn, MoBo gets bigger (it has to house DC-DC converter) and more expensive as well. Currently, price wise, high-end PSU costs about the same as mid-tier MoBo. But in terms of ATX12VO, PSU could cost peanuts while MoBo costs a fortune.

Spoiler: expand for off topic comments

its an example of bad engineering! the PSU is a different kind of engineering from the mobo, the PSU is about converting fixed inputs eg 230volts 50Hz AC, to fixed outputs, eg your 12V DC, whereas the mobo is about totally variable inputs and outputs, namely data. so it is foolish to move the PSU functionality to the mobo, unless you have a really compelling reason.

it just sounds like poor modularity. as a general rule, its best to keep things simple, eg have all the memory sockets in one zone, have all the USB sockets in one zone, have all the power conversion stuff in one separable machine, the PSU.

basically its best to bunch together similar stuff, and where possible make this a separable unit of the system, like the PSU, CPU, mobo, memory.

with programming, its better to have a bit more simplicity at the cost of speed, because the biggest problem is bugs and reading software and improving software leater. so if the software maximises speed but becomes a jungle of activities, it becomes problematic to verify or improve.

one area in bad need of standardisation is smartphone batteries. where each firm seems to have its own batteries, better to standardise like with AA and AAA etc, where duracell, varta, energiser etc can perfect the batteries.

even worse are these unremovable batteries, because when the battery runs out, you have a problem.

laptops with unremovable batteries also are bad, because I like to use my laptop from the mains, eg for lengthy ops eg cloning a disk, this can take hours, and the batteries can run out midway!

but if the battery is permanent, it is continually getting recharged, which cannot be good. why cant they at least put a battery disconnector switch?

* Cable management will go worse as well, since for e.g SATA power (or MOLEX), which has +5V DC and currently comes from PSU, will have connector on the MoBo itself.

At current date, ATX12VO is used with some prebuilt PCs, which already houses proprietary MoBo and PSU. Now, Intel's idea was to streamline power delivery with prebuilt proprietary PCs, but there's no telling that this won't flow over to the current DIY segment of PCs. Heck, you can already buy MoBos that accept only +12V DC.

Intel should stick to designing CPUs! and maybe chipsets,

they are getting ahead of themselves, eg even with Intel CPUs, they try to put stuff to help programmers, but programmers never use this stuff, because programmers typically have to design bespoke protocols to conform with other parts of their work. Intels stuff tends to be at cross purposes.

and in fact, it should be programmers telling Intel what they should supply, not Intel trying to second guess what programmers want!

Further info about ATX12VO if interested:

View: https://www.youtube.com/watch?v=Zc9oRKexV_s



This very chip we talk about, enables MoBo to use USB and PCI-E protocols from the same output port. Normally, USB and PCI-E are separate, with their own separate "data flow" to the system. Now, this specific chip acts as a bridge between USB and PCI-E.

In layman's terms, it enables data from USB port to travel to PCI-E protocol, which is much faster, compared to the USB, where USB data packages have to wait in line.

For analogy example;
Let's say there are two parallel roads going in same direction into the city (into the PC), without ever intersecting with each other.
One road is highway, named PCI-E and has speed limit of 90 KM/h.
Another road is street, named USB and has speed limit of 60 KM/h.

Now, data can travel on both roads, but depending from where they came from, they are either forced to travel on the PCI-E road or on the USB road. Flow on USB road is slower and there can be traffic jams as well due to rush hour.

But if you could build an on-ramp from USB road to PCI-E road, then it alleviates the traffic load on USB road and also would make the traveling data faster, since they can now travel on faster PCI-E road, to the city (into the PC).

ASM3241 chip is that on-ramp between USB and PCI-E, connecting both.

Did you get it now?

ok, I think I understand now, its a kind of USB to PCI-E converter, where it is PCI-E pretending to be USB? where you can attach a USB device, but if you scratch below the surface it is PCI-E?

Here, i need to know the specific GPU you're getting, because i need to look up the ports on GPU. But GPUs can do easy 8K resolution, far more than what MoBo is capable of outputting.

I got this one:

https://www.amazon.co.uk/dp/B0C7WCK...are-gb-1197512415740029200-21&geniuslink=true

There is finite amount of different frequencies a device can use and it is already over saturated.

E.g i can tap into my neighbor's temperature reading unit, since i think we both bought the same Sencor weather station. And i'm sure my neighbor can do the same, reading my weather probes from their device.

in the old days with some TV shops, you could point the same remote control at any TV!

This, of course, brings another issue, if ALL devices are wireless, how can you protect the data when it is carried over air? E.g you are watching something from your wireless monitor, while neighbor also has wireless monitor and can tap in, and also watch the very same thing as you're watching, without you ever realizing it.

you need some kind of encryption.

I think bluetooth "pairs" the devices, where people cant listen in, but I am not an expert!

eg my 3D TV has I think bluetooth active 3D glasses, which you have to pair.

Due to different SSID.

but how does it separate out the wireless of the different modems? these signals will all be mixing in the same space.

eg say if everyone locally uses the identical modem, a bit like with tuning short wave stations where 2 stations could overlap.

the SSID presumes it can, where it can filter out the one with the right SSID, but how does it filter them out to then read off the SSIDs?

does each modem search through a list of available frequencies for a locally unused one?

So, what you're asking is wireless peripherals, that work over wi-fi or bluetooth? E.g wireless mice, wireless KB, wireless headset. But not only those, but wireless monitor as well?

yes! I bought an HDMI transmitter by Maplins, but the problem is it isnt full HD,

I also got a Q-waves one, where there is a slight timelag, where I move the mouse, and a little later the mouse moves.
I think not full HD, I junked them eventually because the audio was kaput. the video transmitted but not the audio.

https://www.amazon.co.uk/Q-Wave-Wireless-USB-HDMI-Extender/dp/B002RL9F38/

its not available as it is obsolete, but you can sometimes get them second hand on ebay.

If so, then monitor takes on the order of up to 100W and would quickly drain any reasonably sized battery. So long as you have to run a power cable to the monitor, there's not much benefit to be had by sending the video over a wireless connection.

Spoiler: these off topic comments are unexpanded to prevent thread proliferation

the products I tried, the monitor or TV uses its own power. there is a benefit, that you dont have a lengthy cable that you have to route around wardrobes and which clutter the floor!

my TV is maybe 4m or 5m from my PC which is in another room. with the above solutions I arranged the video transmitter in this room.

but also the original idea of having a cable free PC, except eg the cables from PSU to mobo. but say no other sockets, everything via some form of wireless or bluetooth or other.

a bit like this insurance replacement vehicle I used once, Nissan X-trail, you just needed to have the key on you, and the vehicle would open and start! but I prefer my more manual controls. That vehicle also gave you an aeriel view of the vehicle for perfect parking manoeuvres, where it must have lots of side cameras, whose images are rejigged with software to form an aeriel view.


also eg you could have a keyboard + monitor which you can take to any room, to any part of the room, without having transport the PC also, and all the associated cables.

data potentially could be transmitted optically, where you have a screen near the ceiling, and a camera at the opposite end of the room reading that signal, a bit like semaphores and smoke signals. remote controls also are an early example of cable free things.


I have a wireless A3 printer, Epson ET-16150, and its great, I have it in one room on some high spacing shelves, and the PC in a totally different room with lack of space. when I print I just walk over to the room with the printer.

potentially 2 computer users can use the same printer also! where its more like a printer server.

I can print on it with my smartphone via the Epson iPrint app.


same way in the old days the modems were all cable based, but at some point they became wireless, where several people can use the same wireless modem via their smartphones wifi or PCs.

so there is significant gain of going from cables to wireless, eg potentially someone can switch their smartphone to the widescreen TV! I think some TVs allow bluetooth.

in an office, you could walk up to a desk, and show a video from your smartphone on the person's monitor!

now my A3 scanner by Plustek is USB cable only, I think USB2, and I have to have that near the PC and this is very awkward, the only option is a long USB cable, but that wont allow the scanner to be put anywhere.

with my car, you can connect the smartphone to the car by bluetooth, where the car becomes a phone!

There are flaws in that idea.

Different PC cases have different amounts of PCI slots. From as little as no PCI slots or two PCI slots, up to 10 or even 12.
Now, are you saying, that MoBo must include separate brackets for: 2 PCI slots, 3 PCI slots, 4 PCI slots, 5 PCI slots, 6 PCI slots, 7 PCI slots, 8 PCI slots, 9 PCI slots, 10 PCI slots, 11 PCI slots, 12 PCI slots? Just to accomodate all PC case sizes? That's 10 different brackets, where user would only use one, depending on how big their PC case is, while the rest, 9 brackets, are completely wasted.

nope! the mobo only supplies a panel for its length plus maybe a small bit alignment to a standard length,

unused zones would be removable panels supplied by the tower. eg my Phantek has 8 removable slots with 7 unremovable separators.

instead it can have 8 removable slots with removable separators, where you can open up variable amounts of the zone.

and have the old era slots with separators if you want.

the mobo can then say arrange 4 slots in the space normally used by 5 to give extra space for the hulking graphics card.

Current design is much better, where there are individual PCI slot covers, whereby one can freely remove only those PCI slot covers they need, while keeping the rest. E.g like i had to do, whereby i removed 2nd and 3rd cover (to house GPU) and also 9th cover, to reroute one cable from inside to outside.

I disagree, because my idea adds to the current design, it doesnt subtract in any way at all, so the current idea cannot be better, as my idea is backwards compatible with the current idea!

the tower will supply those slots you mention, EXCEPT the new standard will demand the separators to be removable, where the entire zone can be opened up, rather than having obstructive separators.

Spoiler: Picture proof

and now show the view of those slots from the outside, I think they will have unremovable separators.

also I am not sure about having a PSU near the floor, as heat rises, so its heat will flow past the mobo, seems better to have the PSU at the top. but the gauze at the top is a good idea allowing air to directly flow out through the top.

with my central heating, the output zone of the fully external flue is ABOVE the air input zone, because heat rises so the input wont draw in the output leading to a carbon monoxide risk.

the idea would need a standardisation of points the mobo panel will reach to, where beyond that you can leave the tower slots and separators in place.

Yes, 1 meter. Min is 30cm. Most SATA data cables are around 50cm.

this does mean if I want to have 3 consecutive cables, I have to go for the 30cm kind of length.

 

[Richard1234]

I finally have all the equipment for the install, although not yet all the sata cables.
in the end I went shopping hoping the amazon cables would arrive, where I went to various shops to use up the time, and they arrived at the collection shop around 7pm.

the torque screwdriver is a digital 1/4" one from CPC Farnell, which is these photos

http://www.directemails.info/tom/torque/sealey1.jpg
http://www.directemails.info/tom/torque/sealey2.jpg
http://www.directemails.info/tom/torque/sealey3.jpg

which is supplied with 2 AAA batteries, and is some £48.89 all inclusive, whereas the 4mm one was some £95, which I have returned for a refund.

for the 1/4" bits, I bought a Rolson for £12, which has precise looking bits:

http://www.directemails.info/tom/torque/rolson4.jpg

and then another Rolson for £10 to get a 1/4" to 4mm converter, which includes some very fine 4mm bits in the handle,

http://www.directemails.info/tom/torque/rolson3.jpg

these 2 from the Asda supermarket DIY aisle, but no such things in other supermarket aisled, nor at B&M,


where the converter info is marked with the green arrow. and for a suite of 4mm bits, the earlier Rolson:

http://www.directemails.info/tom/torque/rolson1.jpg

which is from B&M, where you can see the B&M price of £4.99.

total price: £75.88, whereas with the 4mm torque screwdriver, all I got was the screwdriver for some £95, no bits, and problematic getting a 4mm to 1/4" converter, the only one was a 13 day delay!

obtaining the right torque kit has been the big delaying factor, I could have skipped that and winged it, but as the cost of the system is a bit high, I decided to try and do things correctly, "act in haste, repent at leisure". I didnt get insurance on some of the items, so if I damage them I will be liable!

I wasted some days obtaining the 4mm torque screwdriver, then trying to source bits, then trying to source a 4mm to 1/4" converter. really the MSI instructions should advise on which torque screwdriver to get as its a bit of a jungle, and VERY expensive, £95 screwdriver!

it wasnt a waste of time, I learnt some important stuff trying to go via the 4mm screwdriver.

gonna have supper, and then study Aeacus' instructions again, I might skip the breadboarding!

I need to study the torque screwdriver instructions also! its a bit of a jungle all this, with a manual for the mobo, a manual for the tower, a manual for the screwdriver, a manual for the cooler, etc. I can see why some people buy ready made systems!

I think the mobo manual itself needs a manual, as its a bit of a quagmire

 

[Aeacus]

I have some hardboard, where one side is "rough", see this photo:

http://www.directemails.info/tom/hardboard.jpg

is that static resistant?

what about an ordinary MDF surface?

I dont know if MDF surfaces are a standard material, I just buy the plain ones like in the earlier photo of the table.

I'm not a carpenter, so, i don't know. But internet says MDF is fine regarding ESD,
topic i found: https://www.eevblog.com/forum/reviews/esd-workshop-setup/

I disagree 100%!

Well, then we can agree to disagree.

Random sample of people doesn't accurately show the needs/requirements of the populous. Better sample pool would be where there are people from all walks of life, with different preferences. Since if you can compromise between all of them, you'll end up with a product that is suited for more people; compared to the random people who's ideals mostly may align, but in reality, form a minority. And with this, you'd end up with a product that caters to minority, while the goal was to cater to the majority.

In PC hardware world, what is put out is usually left for consumers to stand up with and get used to. Since if you don't like one brand's product, there is limited option regarding brands. With some components, you only have two choices. E.g CPU, either Intel or AMD. With GPU, you have 3 choices: Intel, AMD or Nvidia. Sure, with GPUs, the selection is wider due to the AIBs, but the core chip and drivers are still from either of the three.

one area in bad need of standardisation is smartphone batteries. where each firm seems to have its own batteries, better to standardise like with AA and AAA etc, where duracell, varta, energiser etc can perfect the batteries.

Spoiler: About replaceable batteries in mobile phones

Mobile phones used to have replaceable batteries. E.g look at the older Nokia phones, e.g infamous Nokia 3310 (dubbed as the most durable mobile phone in the world).

E.g one could easily disassemble entire mobile phone and replace components easily;

I used to use Nokia 3310 a lot in my youth and i also replaced the outer shelling with better eyecandy one (different color and decal). Also, i replaced the keycaps when it wore out. Best part, i could replace battery. Or even carry spare with me, just in case battery got empty and i needed the phone to work.

Another thing going for Nokia 3310 was the insane amount of different covers, where one could customize and personalize their mobile phone for their likening.
E.g just a tiny sample of different cases;

And if you dropped it, at best, casing came off or battery popped out, but after reassemble, it worked without issues.

Today's smart phones are terrible in terms of component replacement. You can't do anything on your own. Only in certified repair shops and even then, not all issues can't be fixed. And when built-in battery dies (or the max capacity dwindles to nothing, as it is with Li-Ion batteries over time), you have to buy a new phone. That's adds a lot to e-waste.

ok, I think I understand now, its a kind of USB to PCI-E converter, where it is PCI-E pretending to be USB? where you can attach a USB device, but if you scratch below the surface it is PCI-E?

Well, it's not PCI-E pretending to be USB, but instead USB being able to send data to PC via PCI-E. And maybe even vice-versa (though, i'm unsure why would you like PCI-E add-in card to output it's data to USB port).

I got this one:

https://www.amazon.co.uk/dp/B0C7WCK...are-gb-1197512415740029200-21&geniuslink=true

Official specs: https://www.msi.com/Graphics-Card/GeForce-RTX-4060-GAMING-X-8G/Specification

Whereby;

Output
DisplayPort x 3 (v1.4a) - it can do 4K at 120 Hz
HDMI™ x 1 (Supports 4K@120Hz HDR and 8K@60Hz HDR and Variable Refresh Rate (VRR) as specified in HDMI™ 2.1a)

Digital Maximum Resolution
7680 x 4320 (8K)

Cables;
2 meters, USB type-C to HDMI 2.1: https://www.amazon.co.uk/UGREEN-Thunderbolt-Compatible-MacBook-Surface/dp/B0BXWBSJ77/
3 meters, USB type-C to DP 1.4: https://www.amazon.co.uk/UGREEN-DisplayPort-Thunderbolt-Compatible-MacBook/dp/B0CDW73QH7

With DP cable, there are 3 length options: 1, 2, and 3 meters.

I think bluetooth "pairs" the devices, where people cant listen in, but I am not an expert!

With wireless data transfer, you can always listen in. It is then just the matter of decoding the data flow to readable info.

For example, when i pair my smart watch with my smart phone, i can scan all the bluetooth signals my phone picks up, and it has picked up 2nd phone, which isn't from my household (most likely my neighbors, though the wall). I've also picked up LG TV, which is funny since i have Sony and Samsung TVs in my home. The SSID shows the name, so i can tell what device my phone found.

but how does it separate out the wireless of the different modems? these signals will all be mixing in the same space.

eg say if everyone locally uses the identical modem, a bit like with tuning short wave stations where 2 stations could overlap.

the SSID presumes it can, where it can filter out the one with the right SSID, but how does it filter them out to then read off the SSIDs?

does each modem search through a list of available frequencies for a locally unused one?

SSID, besides being a name of the network, is also unique ID. Due to that, several modems can use the same frequency but since each and every one has unique ID, PC can tell a diff between them.

Now, i'm not a network expert and can't quite explain it in layman's terms. Maybe someone else comes along and explains it better.
In the mean time, found this explanation: https://superuser.com/questions/124...e-difference-between-one-computer-and-another

I also got a Q-waves one, where there is a slight timelag, where I move the mouse, and a little later the mouse moves.

Every wireless connection has worse latency compared to the wired connection. This is one of the reasons why i prefer wired connections.

but also the original idea of having a cable free PC, except eg the cables from PSU to mobo. but say no other sockets, everything via some form of wireless or bluetooth or other.

Spoiler: More about wireless devices

I get that you want convenience, but with your gained convenience, criminals also gain convenience.

E.g keyless cars are one such thing, that i will NEVER get for myself, since it is very easy for criminals to copy and mirror the signal, thus stealing the car under 60 seconds, without a scratch on the car.

Prime example from UK;

View: https://www.youtube.com/watch?v=bR8RrmEizVg


That Mercedes easily costs £100.000 and just look how easily it is stolen. No lockpicking on the door, no lockpicking the ignition. Just mirror the keyless signal, start the engine, sit in and drive away.

and now show the view of those slots from the outside, I think they will have unremovable separators.

Yes, my PC case does have separators between PCI slots. It's part of the PC case structural integrity.

the tower will supply those slots you mention, EXCEPT the new standard will demand the separators to be removable, where the entire zone can be opened up, rather than having obstructive separators.

Rather than demanding it and forcing your idea on to everybody, you could then just buy that niche PC case that has open area around PCI slots.

E.g Antec Performance 1 FT;
specs: https://www.antec.com/product/case/performance-1-ft

Then Fractal Design have list of PC cases that have bridgeless (open) PCI slots.
Full list here: https://support.fractal-design.com/...acket-compatibility-with-fractal-design-cases

And so does every "open air" PC case as well. E.g Thermaltake offers plenty of "open air" style PC cases,
lineup: https://www.thermaltake.com/products/chassis.html?cat=190&product_list_limit=30

also I am not sure about having a PSU near the floor, as heat rises, so its heat will flow past the mobo, seems better to have the PSU at the top. but the gauze at the top is a good idea allowing air to directly flow out through the top.

PSU is closed on top, whereby when installing it at the bottom, means PSU fan is facing downwards. This way, PSU has it's own airflow path, completely separate from the rest of the system.

This is the old design, with top mounted PSU. In this case, PSU has to take the hot air from inside the case, to cool itself. Since air inside the PC case is already heated up by the CPU and GPU (if using air cooler on both), PSU has only hot air to take, in effort to cool itself. If it even can. This leads to PSU fan spinning faster and noisier, since intake air is hot and won't cool the PSU as well as cold air would.

This is the new design, with bottom mounted PSU. In this case, PSU has easy access to the cold air from outside, from the bottom of the PC case, to draw in and cool itself. This leads to slower fan RPM in PSU and in turn, also less noise from PSU. Moreover, PSU has better cooling due to the cold air it can take directly from outside.

With top mounted PSU, you have no choice, you have to install PSU with fan facing downwards (some few cases may allow installing fan facing upwards, if there is grille at the top).
But with bottom mounted PSU, you have an option, either to mount the PSU fan facing downwards, or upwards (that is, if your PC case doesn't have closed PSU shroud without grille on top).

this does mean if I want to have 3 consecutive cables, I have to go for the 30cm kind of length.

Yes.

I might skip the breadboarding!

It's your funeral. 🪦

 

[35below0]

There are a few cases with the PSU on top. The ones i have seen install the PSU upside down so the fan faces out and up.
The design trades up space at the bottom of the case for space at the top. It's... an idea i suppose. It limits water cooling radiator options somewhat but it's a workable solution in some circumstances.

As for breadboarding. It's skippable. I've skipped it when i built my recent PC. But consider what it is you are skipping. It's an opportunity to catch problems early. Electronics usually work ok for the duration of their warranty or longer, IF they are not low quality junk. But sometime even premium components are dead on arrival or show signs of problems.

Without breadboarding you hook everything up, stuff it in the case, cross your fingers it will be ok. If it's ok, no problem. If it's not all ok... You have to take everything out and disassemble it in order to trace the problem.

If you're affraid of it, that's understandable. And it will probably be fine. It's your own blank cheque you're writing