SV266A upgrade?

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Looking for a budget upgrade. Current system is an Intel CA810e with Celeron
900MHz and 256MB SDRAM. Also in the box: PCI RAVE TNT 16MB, PCI Ethernet
10/100 card, PCI 56k internal modem card, PCI 4 port USB 1.0 card, ST340016A
7200rpm 40GB HDD, FDD, nu DDW-081 DVD-RW drive loaded with XP Pro all in a
200W ATX case (though I have a 250W available also).

Was considering buying Syntax SV266A from eBuyer (cheapest?) and a Duron
1800 MHz plus HSF.

Questions:

- what's the minimum power PSU recommended?
- anyone know the difference between eBuyer's OEM & Retail versions? I/O
shield?
- any problems flashing the BIOS; some think you need a slow Duron to do
this?
- how much faster is this likely to be with the SDRAM in it? (I know DDR
will be faster)
- I'll be losing my gameport/MIDI as this mobo doesn't have it; any
suggestions on a replacement?
- is XP likely to boot off the existing drive with the new mobo/CPU?

Thanks

Tim

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Wed, 5 May 2004 22:20:35 +0100, "Tim Steele" <tjfs@nospamtjfs.com>
wrote:

>Looking for a budget upgrade. Current system is an Intel CA810e with Celeron
>900MHz and 256MB SDRAM. Also in the box: PCI RAVE TNT 16MB, PCI Ethernet
>10/100 card, PCI 56k internal modem card, PCI 4 port USB 1.0 card, ST340016A
>7200rpm 40GB HDD, FDD, nu DDW-081 DVD-RW drive loaded with XP Pro all in a
>200W ATX case (though I have a 250W available also).
>
>Was considering buying Syntax SV266A from eBuyer (cheapest?) and a Duron
>1800 MHz plus HSF.

Buying the cheapest might be a mistake, and this late into the socket A
era there's little point buying such an old low-feature chipset or board.
For better features and upgradability I suggest buying a KT600 chipset
board, or for a few $ more, an nForce2-400 chipset.

>Questions:
>
>- what's the minimum power PSU recommended?

Name-brand 300W. For example, Antec, Sparkle, Fortron, Delta, PC Power &
Cooling.

>- anyone know the difference between eBuyer's OEM & Retail versions? I/O
>shield?

Providing it fits the motherboard (???) the difference shouldn't matter.
Generally a retail motherboard comes with an I/O shield but (especially if
it has the standard port arranagement (old style with analog joystick
port, 3 audio jacks under it, and only 2 USB ports on the rear) any
generic shield with this now-aging standard port cutout should work,
providing your system case doesn't have an unusual method of mounting that
I/O shield.


>- any problems flashing the BIOS; some think you need a slow Duron to do
>this?

Generally no, it should be straightforward and easy, though with that
cheap board there "might" be more potential for problems, but even so
flashing the bios is pretty straightforward. On the other hand, if you
mean that the board doesn't natively support the Duron 1.8, I can only
speculate that it should work, that you wouldn't need a slower CPU.

>- how much faster is this likely to be with the SDRAM in it? (I know DDR
>will be faster)
>- I'll be losing my gameport/MIDI as this mobo doesn't have it; any
>suggestions on a replacement?
>- is XP likely to boot off the existing drive with the new mobo/CPU?

Faster than what, the Celeron 900? Depends on where the bottleneck is in
each app. For websurfing and general office use you probably won't notice
much difference, since you're keeping same HDD and Celeron 900 is fast
enough for those uses. For gaming it should be significantly faster but
the PC-133 memory will be more and more of a bottleneck.

XP will not be likely to boot, you'd need to do a repair install before
trying to boot the new board, then reupdate the patches and such.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Wed, 05 May 2004 22:20:35 +0100, Tim Steele wrote:

> Looking for a budget upgrade. Current system is an Intel CA810e with Celeron
> 900MHz and 256MB SDRAM. Also in the box: PCI RAVE TNT 16MB, PCI Ethernet
> 10/100 card, PCI 56k internal modem card, PCI 4 port USB 1.0 card, ST340016A
> 7200rpm 40GB HDD, FDD, nu DDW-081 DVD-RW drive loaded with XP Pro all in a
> 200W ATX case (though I have a 250W available also).
>
> Was considering buying Syntax SV266A from eBuyer (cheapest?) and a Duron
> 1800 MHz plus HSF.
>
> Questions:
>
> - what's the minimum power PSU recommended?

When you can get 500W for about the same price as 300W (recommended
minimum), what does it matter. Get the 500W. You won't be sorry.

> - anyone know the difference between eBuyer's OEM & Retail versions? I/O
> shield?

I don't even know or care who ebuyer is.

> - any problems flashing the BIOS; some think you need a slow Duron to do
> this?

If there's a problem flashing the bios on the board, don't get it. I
wouldn't get that old board anyway.

> - how much faster is this likely to be with the SDRAM in it? (I know DDR
> will be faster)

More than twice as fast, even with that old board.

> - I'll be losing my gameport/MIDI as this mobo doesn't have it; any
> suggestions on a replacement?

Yeah, I suggest a different board all together. And unless you have PC133
ram in you celery system you'll need to buy new ram anyway.

> - is XP likely to boot off the existing drive with the new mobo/CPU?
>
Nope. You'll need to do a repair install I'm told. I don't use Win.

The 1800 Duron is likely to be a Tbred B core (CPUID 681). It's default
multiplier will be 13.5 and the default FSB will be 133MHz for 1800MHz
core speed. This little cpu will probably run up close to 2400MHz. But
it's probably going to be multiplier locked, so you can only raise the FSB
to increase the speed. The SV266A won't go much higher than about 145, if
that high. And even with PC133 ram, you may have a problem with that.
There's plenty of cheap MB's that will run 200MHz FSB and over, so why
waste time with that old board?
Just my 2 cents. Well, maybe a dime.

--
Abit KT7-Raid (KT133) Tbred B core CPU @2400MHz (24x100FSB)
http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Thu, 06 May 2004 07:09:57 GMT, Wes Newell <w.newell@TAKEOUTverizon.net>
wrote:

>On Wed, 05 May 2004 22:20:35 +0100, Tim Steele wrote:
>
>> Looking for a budget upgrade. Current system is an Intel CA810e with Celeron
>> 900MHz and 256MB SDRAM. Also in the box: PCI RAVE TNT 16MB, PCI Ethernet
>> 10/100 card, PCI 56k internal modem card, PCI 4 port USB 1.0 card, ST340016A
>> 7200rpm 40GB HDD, FDD, nu DDW-081 DVD-RW drive loaded with XP Pro all in a
>> 200W ATX case (though I have a 250W available also).
>>
>> Was considering buying Syntax SV266A from eBuyer (cheapest?) and a Duron
>> 1800 MHz plus HSF.
>>
>> Questions:
>>
>> - what's the minimum power PSU recommended?
>
>When you can get 500W for about the same price as 300W (recommended
>minimum), what does it matter. Get the 500W. You won't be sorry.

But you can't. You could get a 300W-true-capacity generic which has been
labeled as a 500W, being only as well off or worse than a decent 300W.
There are no 500W units worth anywhere near 500W for under $60. A
name-brand 300W can be had for $35. That's not "about the same price".

Pick a generic 500W and put it to the test. "It powers system A" is not a
test, a test is putting it at load rated on label or at very least a proof
that it would power anything more, be better in any way than a 300W
name-brand. By name-rand I mean a power supply manufacturer's name on the
label, not just a more popular relabeled generic.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Thu, 06 May 2004 07:47:58 +0000, kony wrote:

> On Thu, 06 May 2004 07:09:57 GMT, Wes Newell <w.newell@TAKEOUTverizon.net>
> wrote:
>>When you can get 500W for about the same price as 300W (recommended
>>minimum), what does it matter. Get the 500W. You won't be sorry.
>
> But you can't. You could get a 300W-true-capacity generic which has been
> labeled as a 500W, being only as well off or worse than a decent 300W.
> There are no 500W units worth anywhere near 500W for under $60. A
> name-brand 300W can be had for $35. That's not "about the same price".
>
So, all the approvals from American, European. Asian, and other countries
around the world that have certified these PSU's mean nothing? There's
really not much to a PSU. The cost is about 10% material and 90% labor.
Some places labor is less than one US dollar a day. Of course a US
manufactured PSU will have labor cost of about $100 a day or more. So if
you think the cost really has anything to do with it you're dreaming.

> Pick a generic 500W and put it to the test. "It powers system A" is not
> a test, a test is putting it at load rated on label or at very least a
> proof that it would power anything more, be better in any way than a
> 300W name-brand. By name-rand I mean a power supply manufacturer's name
> on the label, not just a more popular relabeled generic.

All my cheap PSU's have name brands on the label, along with
specifications and certifications from at least 5 certifying countries.
Name does not assure quality. And lack of a name doesn't mean it's any
less comparable to a name brand.
The PSU's I've used for AMD 64 system;
400W Powerline. Works well with AMD XP system. Won't power AMD64
550W Power Magic, No problem.
500W Lead Power, ditto
600W Lead Power, ditto

The Powerline had a multiple unit spec label on it for different models
and a generic 400W sticker. This one I think is mislabeled. The others
only have specs for their model which matched the box label. The most
expensive was the 600W ($24) and the least was the 550W I think. it was
$15 with free shipping, which bought the cost to probably $11. The 400W
that didn't work with the AMD 64 came in a bundle and i don't know it's
cost, although I only paid $100 for case/PSU/K7S8X MB/1800+CPU/CPU cooler.

Now, do you really think it would concern me if they only performed to
90% of there rated value.


--
Abit KT7-Raid (KT133) Tbred B core CPU @2400MHz (24x100FSB)
http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 07 May 2004 03:49:01 +0000, kony wrote:

> Without testing you are only guessing that those units would run at 90% of
> rated capacity. They're inferior to many name-brand units of 300W, and
> not are the equal of a most any name-brand 400-420W unit. If you only
> need 300W there's no point in buying a generic with other reliability and
> noise, surge protection circuits missing just to end up with same or lower
> true output. Your $ is supporting fraud when you buy those generics.

I don't know who you are trying to convince, but it sure won't be me.

--
Abit KT7-Raid (KT133) Tbred B core CPU @2400MHz (24x100FSB)
http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 07 May 2004 09:14:08 GMT, Wes Newell <w.newell@TAKEOUTverizon.net>
wrote:

>On Fri, 07 May 2004 03:49:01 +0000, kony wrote:
>
>> Without testing you are only guessing that those units would run at 90% of
>> rated capacity. They're inferior to many name-brand units of 300W, and
>> not are the equal of a most any name-brand 400-420W unit. If you only
>> need 300W there's no point in buying a generic with other reliability and
>> noise, surge protection circuits missing just to end up with same or lower
>> true output. Your $ is supporting fraud when you buy those generics.
>
>I don't know who you are trying to convince, but it sure won't be me.

It's up to anyone to use the facts any way they like. The fact is that
the power supplies you mentioned aren't worth anywhere near their rated
output and trying to use one at that rated output is asking for trouble.
It's your money and your gamble, but even OEMs don't build such a low
quality per watt into their units and certainly they are keeping an eye on
design costs.

If you didn't need to realize the false output rating then there was no
reason to choose the generic in the first place, a name-brand of same true
output is a very similar cost but still usually better quality. The power
supply industry is competitive, except for eye-candy brands like Enermax
the majority of the producers have their eye on OEM sales and produce very
good value units, it's not just a markup for the name on a "name-brand",
with the notable exception of PC Power & Cooling, but then they have
distinction by "many" to be the best, and so it goes that the best usually
carries a disproportionately higher pricetag.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 07 May 2004 10:50:11 +0000, kony wrote:

> On Fri, 07 May 2004 09:14:08 GMT, Wes Newell <w.newell@TAKEOUTverizon.net>
> wrote:
>
>>On Fri, 07 May 2004 03:49:01 +0000, kony wrote:
>>
>>> Without testing you are only guessing that those units would run at 90% of
>>> rated capacity. They're inferior to many name-brand units of 300W, and
>>> not are the equal of a most any name-brand 400-420W unit. If you only
>>> need 300W there's no point in buying a generic with other reliability and
>>> noise, surge protection circuits missing just to end up with same or lower
>>> true output. Your $ is supporting fraud when you buy those generics.
>>
>>I don't know who you are trying to convince, but it sure won't be me.
>
> It's up to anyone to use the facts any way they like. The fact is that
> the power supplies you mentioned aren't worth anywhere near their rated
> output and trying to use one at that rated output is asking for trouble.

Sorry, but you just don't know wtf you are talking about.

> It's your money and your gamble, but even OEMs don't build such a low
> quality per watt into their units and certainly they are keeping an eye on
> design costs.
>
And I've gambled over 400 times and won then.

> If you didn't need to realize the false output rating then there was no
> reason to choose the generic in the first place, a name-brand of same
> true output is a very similar cost but still usually better quality.

Well, I'm not sure what you mean by name brand at this point. Are these
name brand?
Lead power
Power magic by Xcase

And are you stating that they don't come anywhere near their rating, and
what do you consider "anywhere near"?

--
Abit KT7-Raid (KT133) Tbred B core CPU @2400MHz (24x100FSB)
http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 07 May 2004 22:08:43 GMT, Wes Newell <w.newell@TAKEOUTverizon.net>
wrote:

>On Fri, 07 May 2004 10:50:11 +0000, kony wrote:
>
>> On Fri, 07 May 2004 09:14:08 GMT, Wes Newell <w.newell@TAKEOUTverizon.net>
>> wrote:
>>
>>>On Fri, 07 May 2004 03:49:01 +0000, kony wrote:
>>>
>>>> Without testing you are only guessing that those units would run at 90% of
>>>> rated capacity. They're inferior to many name-brand units of 300W, and
>>>> not are the equal of a most any name-brand 400-420W unit. If you only
>>>> need 300W there's no point in buying a generic with other reliability and
>>>> noise, surge protection circuits missing just to end up with same or lower
>>>> true output. Your $ is supporting fraud when you buy those generics.
>>>
>>>I don't know who you are trying to convince, but it sure won't be me.
>>
>> It's up to anyone to use the facts any way they like. The fact is that
>> the power supplies you mentioned aren't worth anywhere near their rated
>> output and trying to use one at that rated output is asking for trouble.
>
>Sorry, but you just don't know wtf you are talking about.

Don't be sorry, TEST those generics. Open them, inspect them, pull the
spec sheets from the components contained within and compare them to any
number of the more popular name-brand units. No amount of attitude
replaces fact.

>
>> It's your money and your gamble, but even OEMs don't build such a low
>> quality per watt into their units and certainly they are keeping an eye on
>> design costs.
>>
>And I've gambled over 400 times and won then.

More likely you never used them in systems that approached needing 500W,
which was half of my argument all along.


>
>> If you didn't need to realize the false output rating then there was no
>> reason to choose the generic in the first place, a name-brand of same
>> true output is a very similar cost but still usually better quality.
>
>Well, I'm not sure what you mean by name brand at this point. Are these
>name brand?
>Lead power
>Power magic by Xcase

Manufacturer name brand. No, those aren't name-brand, merely some of the
more popular generic labels.

>
>And are you stating that they don't come anywhere near their rating, and
>what do you consider "anywhere near"?

Typical name-brand units are spec'd at 20-45C ambient, 100K (or only >=50K
if employing a sleeve bearing fan) MTBF, and max _sustainable_, continuous
output. This does not include allowance for periodic surge ability... a
decent name-brand labeled as 300W may easily be able to output upwards of
400W during peak loads, for example, spinning up HDDs during initial
power-on of system. Because output power, temp, and lifespan are related,
those specs must remain constant for a comparion. Given those constants,
the generics (which in this case, those you mention appear to be relabeled
CWT low-end parts) aren't capable of 80% of their rated wattage, as a very
rough figure... that's the thing with generics, the label can and is
altered and so a continuous offset like "80%" cannot be assumed. For
example, I had a stack of old generics thrown out during spring cleaning
that were labeled as 200-250W units. I still have one that's identical
inside but labeled as 350W. The difference is that it was sold more
recently and label upgraded to reflect the market's demand for higher
wattage units, but a demand which in this case was never met by even the
slightest attempt. Any units that have a 200W label are certainly rated
more accurately, less subject to an 80% power factor than those labeled as
350W.