PSU tier list 2.0

Page 170 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.

Seriously, you're taking that line?

Anyway, no. Running RAM above 1.5V on Intel DDR3 memory controllers is out-of-spec and may cause long-term degradation. Just as dropping to 10.8V before making PWR_OK inactive is technically out of spec, but may or may not actually cause issues.

Using a 28-lane LGA2011-3 CPU will mean that you can't take full advantage of your motherboard, particularly with regards to multiple GPUs or M.2 drives - it "works", but not as it should.

Intel Quicksync only works under certain circumstances, and its capabilities depend on loads of factors.

Sandy Bridge's integrated graphics can only drive two displays - but you'll never find this info in the motherboard specs.
 


Yes, I am taking that line. And yes all that stuff is true.

Just above, Karadjgne said UVP and OVP monitored input voltages. But you did not correct him. I believe that if I had said that, you would have corrected me. Which I'm cool with of course.
 
Yep, I was tired when I posted that, actually beyond tired, and while the answer was accurate for what I deal with, it was dead backwards for a psu. Trust me when I tell ya that ppl like SS, onus, and ehem.... darkbreeze don't miss much. I'm actually a little surprised, musta caught darkbreeze napping. I think db must get a perverse sorta pleasure out of saying things to me in the nicest possible way while making me feel like a total idiot at the same time.

Don't feel bad turkey, you aren't alone at the bottom of the totem pole.
 
It's just hard. I try to understand something, but the description of that something contains too much other stuff within which I don't even know, and if I try to understand something of that, same thing.

I guess the trail all leads back down to charge and fields? If you keep asking how stuff works, then ask how the stuff that makes that something work works, it'll eventually root all the way down to atoms.
 

That could be said for literally everything.

I found that turning all the links purple on Wikipedia helps quite a lot.
 
On Hardwaresecrets reviews, could someone explain to me what Torres means by:
In the switching section, this power supply uses two SPW32N50C3 MOSFETs in a resonant configuration. Each transistor supports up to 32 A at 25º C or 20 A at 100º C in continuous mode, or up to 96 A at 25º C in pulse mode, with a maximum RDS(on) of 110 mΩ.
Source: http://www.hardwaresecrets.com/fractal-design-newton-r3-800-w-power-supply-review/4/

1) Continuous vs pulse mode, what is the difference?
2) These transistors support a certain limit of current based on the temperature. What, though, in this case, would be the power they support? What is the voltage?
3) RDS on? What is that?

I've read a lot of Hardwaresecrets reviews, but I've always sort of gleamed past this stuff, but I'd like to know more about exactly what it means.
 
1) Continuous means you're leaving the transistor turned on for long enough periods of time that it reaches a steady state - say, using it to switch a fan or light on/off. Pulsed would mean you're switching it continuously; e.g. PWM or driving a switchmode supply. This number is *very* fluffy, and depends on loads of assumptions like duty cycle, frequency, voltage, what type of load etc. Read the datasheet.

2) Consider reading the datasheet, but I think it depends a lot on how they're being used.

3) RDS(on) means Resistance, Drain to Source (D&S should be in subscript), while the transistor is switched on - google would have found oh so many copies of the same answer. Lower is better, as you get less heat and less voltage drop. Note that it's not the be all and end all - gate capacitance can generate loads of heat at higher switching frequencies.

Read the datasheet is the answer to pretty much any question about specific components.
 
Tell me, why do we have teachers? If all knowledge already exists in books, PDF documents, and web pages, what need is there for any teachers in this world? The answer is that a teacher is better. A teacher is always better than a datasheet. It's because a teacher has the ability to communicate with the student in a relatable manner. I'll never give in to the concept that all knowledge can be obtained without person-to-person communication. A document written to an arbitrary reader may not relate very well to that reader or to that reader's current level of knowledge.

That data sheet is 20 pages of columns and rows of stuff I don't know. How can you expect me to comprehend it even remotely when I know only the basics of SMPS? Throughout all my schooling, a teacher has always been better than a book, and it remains so that, while Google may have the answer to everything, Google may not be able to present that information in a comprehensible manner to a wide range of readers, including me.
 
A 'teacher' is not always better. I am reciting the above purely from memory, and there are almost certainly inaccuracies, and it may well not apply to the case in question. Datasheets are written by people with lots of experience of the part in question, whereas I've never heard of it. Plus I'm only one (entirely fallible) source.

Also, you will find lots of better explanations on Google than I can provide (particularly StackExchange, who place a lot of emphasis on an answer being useful to not just the original questioner), including 'how to interpret a datasheet'. They are far more than just columns of raw data, and often include application notes.

Expecting someone to re-create material that already exists, just for you, is little more than selfish.

How can you expect me to comprehend it even remotely when I know only the basics of SMPS?
Google, largely. Where do you expect I found anything. Google, and think about whether anything contradicts anything you've heard before, or seems blatantly illogical. Also, watch a bunch of videos of people reverse-engineering basic goods (cheap stuff out of china tends to be good for laughs) - they tend to throw in a lot of comments of why people would do stuff.

Switchmode power supplies are fairly advanced circuits; you should probably look into understanding some of the basic characteristics of components like FETs if you ever want to understand anything other than by rule of thumb.
 
I was under the impression that you have like 40 years of electrician work behind your back and have an EE degree or something... I'll look into FETs. Beaty had a great transistor explanation that made sense actually, every other one didn't make sense to me.
 


Huh I thought "senior moderator" literally meant senior. Strange, I always pictured you in your 60s.

See, it seems like most people have fathers with background in this stuff except me. My family is all law and business 😛
 


Still, you're one of the smartest 19 year old members we or any other forum has. You should keep in mind that not everybody has your aptitude or capacity to absorb information. Takes some people longer than others to learn concepts, but that doesn't mean they're not incapable of learning them to one degree or another.
 
Sorry; I probably over-reacted a bit there.

It feels a little wrong to me to ask a question about something without at least trying to look on El Google, and it's something that I see lamented a lot here, though most of our members at least have the excuse of needing a personalised answer ("I need a new GPU, to go with this near-unique set of other components").

If you need a page worth of explanations to understand the answer, what makes you think a 'teacher' can do it in a sentence?
 
I didn't think I needed a page worth of explanations...

Okay, let me think about all the questions I have, and I'll Google one of them. I'll Google "In a power supply, what is the return path for charge?"

Since charge in a circuit must go around the circuit, there much be a return path. If there's a 12V wire going to a GPU, how can that charge flow in just that one direction without flowing back via a second wire? I'm guessing that is the ground wire, but let's see if my prediction is correct.

Results: https://www.google.com/search?q=In+a+power+supply%2C+w...

Hey, my prediction was actually correct! But that brings another question. What if the outlet is not grounded? I'll search "if an outlet is not grounded how do ground wires provide a return path?".

Results: https://www.google.com/webhp?sourceid=chrome-instant&ion=1&espv=2&ie=UTF-8#q=if+an+outlet+is+not+grounded+how+do+ground+wires+provide+a+return+path%3F

Hmm, this is not what I was looking for. These are more pertaining to what a ground wire in general is rather than answering my particular question... This is when I come to a forum. But what if the people on the forum tell me to Google it? How will I ever learn?
 
Current returns to the source, not to ground. If current went out of the PSU via the supply ground wire, how would it return to the PSU?

I'm not saying that you can't ask questions - but try to google them first. You might also find that an electronics specific forum can be more helpful than us on those types of questions.
 
To correctly understand the answer, you'll have to learn the correct terminology. In AC power there are 3 electrical paths, hot-neutral-ground. In DC there are 3 also, positive-negative-ground. Negative and neutral are sometimes confused with ground and sometimes used by grounding. In a psu, after the AC-DC conversion, there should be nowhere that negative touches ground. Ground serves 2 purposes, first is safety, a pathway back to source in the advent of component failure, and the second is antenna. Any and all outside rf /static frequencies that would induce/interfere with the workings of the psu are shunted harmlessly away. Negative is different, that return path carries the unbalanced load, the leftovers from what's actually used. A psu will work ungrounded just fine but not for long. What'll invariably happen is static charge will build and have no outlet until it creates one somewhere via short to a powered component where the differential in polarities will act as an attraction, like 2 magnets. Bye bye motherboard
 
Status
Not open for further replies.