antec 550w w/ LOW +5v rail

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

i just noticed my +5v rail is running between 4.38 to 4.5 volts. not
sure how long it has been doing this, but my system has been a little
erratic over the last few weeks. normally my rig is rock solid. i
dropped my fsb from 200 back to 166, lowered my voltages back to
normal, disconnected all of my drives, and it is still on the low
side. i've had my antec true power 550w for about 2 years (maybe a
little longer). i thought i had a bad sata hdd or dvd burner, but
nogo. also my 3.3 is about 3.28, and my 12 is 11.37. my system and my
22"nec monitor are plugged into an APC Back-UPS Pro 1000. does this
sound like an RMA issue? seems like it to me. thanks in advance.

System Specs:
Asus A7N8X Del 2.0 (1007 Uber bios)
2500+ @3200+ (2.2 Ghz) 200 fsb
Thermalright SLK-800(A) (with AS3 and 80mm Denki fan)
1Gb Corsair pc3200 CL2 (2 sticks, with platnium heat spreaders)
Radeon 9700 Pro retail w/Aluminum ram sinks
22in NEC MultiSync FE1250
160 gig Maxtor SATA 7200rpm 8mb (XP Home SP1)
120 gig Maxtor PATA 7200rpm 8mb (data/backups)
Optorite dd0203 DVD burner
Artec 52x24x52 cdrw
Lian-Li pc10 (with window and light mod)
550W Antec True Power PSU
Phoebe v.92 pci modem
onboard sound
onboard nic

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Il Wed, 09 Jun 2004 04:30:37 -0500, OverKlocker ha scritto:

> i just noticed my +5v rail is running between 4.38 to 4.5 volts.

Yes, that is the problem of your system instability.

Definitely broken PSU, I think. Or do u maybe have electrical problem in
the house?

 

[Nero]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Mine is at 4.84V and I have no problems.
Is that reading taken accross the - and + 5Volt rail or just the reading
your software tells you?
The only way to take voltage readings is by connecting a volt/test meter
directly to the leads/connectors on the PSU and not any other way.
What you see is after the MB regulates the voltage and not off the PSU.

 

[Nero]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

You will only get a proper reading by connecting a test/volt meter across
the - and + 5 volt rail on the psu.
You cannot get an accurate reading by software or by what it says anywhere
else.
Mine is 4.84V and never any problems or slowing down.
The reading you see via windows or software is after the onboard regulation.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

On Wed, 09 Jun 2004 12:30:39 +0200, _P_e_ar_lALegend <end_6@leg.it>
wrote:

>Il Wed, 09 Jun 2004 04:30:37 -0500, OverKlocker ha scritto:
>
>> i just noticed my +5v rail is running between 4.38 to 4.5 volts.
>
>Yes, that is the problem of your system instability.
>
>Definitely broken PSU, I think. Or do u maybe have electrical problem in
>the house?

power in the house is fine, plus my APC Back-UPS Pro 1000 should
condition the power if it was off slightly. more than likely, i'm will
be calling antec...LOL

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Il Wed, 09 Jun 2004 13:07:41 +0100, Nero ha scritto:

> You will only get a proper reading by connecting a test/volt meter across
> the - and + 5 volt rail on the psu.

Well, hardware monitor, especially the one implemented in the bios, is not
perfect, but quietly accurated. If it say 4,3 and u have problem, it's
becouse it really is 4,3.

> Mine is 4.84V and never any problems or slowing down.

4.84 is acceptable. 4.5 or less it's not.

the +5V is the most important reading for the stability and the
performance of the system: it handle the cpu power and have to be
perfectly stable and power enough.

My Enermax is actually 5.01 and system run great.

I had, in the past, low cost power supply with 4.7/4.8 reading and system
never been so stable like it's now.

I can say it's much more better to have a HIGH quality 250 PSU (I do have
an old 250 AOpen, rock steady at 5,03 running a totally stable 333/2400+
on an Abit board).

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

Your power supply is failing.

--
DaveW



"OverKlocker" <nunya@biznez.net> wrote in message
news:79ldc0pomrbc1lovvl4r88guqcfhl70k06@4ax.com...
> i just noticed my +5v rail is running between 4.38 to 4.5 volts. not
> sure how long it has been doing this, but my system has been a little
> erratic over the last few weeks. normally my rig is rock solid. i
> dropped my fsb from 200 back to 166, lowered my voltages back to
> normal, disconnected all of my drives, and it is still on the low
> side. i've had my antec true power 550w for about 2 years (maybe a
> little longer). i thought i had a bad sata hdd or dvd burner, but
> nogo. also my 3.3 is about 3.28, and my 12 is 11.37. my system and my
> 22"nec monitor are plugged into an APC Back-UPS Pro 1000. does this
> sound like an RMA issue? seems like it to me. thanks in advance.
>
> System Specs:
> Asus A7N8X Del 2.0 (1007 Uber bios)
> 2500+ @3200+ (2.2 Ghz) 200 fsb
> Thermalright SLK-800(A) (with AS3 and 80mm Denki fan)
> 1Gb Corsair pc3200 CL2 (2 sticks, with platnium heat spreaders)
> Radeon 9700 Pro retail w/Aluminum ram sinks
> 22in NEC MultiSync FE1250
> 160 gig Maxtor SATA 7200rpm 8mb (XP Home SP1)
> 120 gig Maxtor PATA 7200rpm 8mb (data/backups)
> Optorite dd0203 DVD burner
> Artec 52x24x52 cdrw
> Lian-Li pc10 (with window and light mod)
> 550W Antec True Power PSU
> Phoebe v.92 pci modem
> onboard sound
> onboard nic
>

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

My 550W enermax gave some very strange voltages (including low outputs) when
the +12V supply was not actually loaded hard enough.
I noticed that when I added more disk drives, the voltages became perfect.
If you read the specifications carefully, most power supplies have a minimum
load requirement on some of the outputs.
Just a thought....

- Steve

"OverKlocker" <nunya@biznez.net> wrote in message
news:79ldc0pomrbc1lovvl4r88guqcfhl70k06@4ax.com...
> i just noticed my +5v rail is running between 4.38 to 4.5 volts. not
> sure how long it has been doing this, but my system has been a little
> erratic over the last few weeks. normally my rig is rock solid. i
> dropped my fsb from 200 back to 166, lowered my voltages back to
> normal, disconnected all of my drives, and it is still on the low
> side. i've had my antec true power 550w for about 2 years (maybe a
> little longer). i thought i had a bad sata hdd or dvd burner, but
> nogo. also my 3.3 is about 3.28, and my 12 is 11.37. my system and my
> 22"nec monitor are plugged into an APC Back-UPS Pro 1000. does this
> sound like an RMA issue? seems like it to me. thanks in advance.
>
> System Specs:
> Asus A7N8X Del 2.0 (1007 Uber bios)
> 2500+ @3200+ (2.2 Ghz) 200 fsb
> Thermalright SLK-800(A) (with AS3 and 80mm Denki fan)
> 1Gb Corsair pc3200 CL2 (2 sticks, with platnium heat spreaders)
> Radeon 9700 Pro retail w/Aluminum ram sinks
> 22in NEC MultiSync FE1250
> 160 gig Maxtor SATA 7200rpm 8mb (XP Home SP1)
> 120 gig Maxtor PATA 7200rpm 8mb (data/backups)
> Optorite dd0203 DVD burner
> Artec 52x24x52 cdrw
> Lian-Li pc10 (with window and light mod)
> 550W Antec True Power PSU
> Phoebe v.92 pci modem
> onboard sound
> onboard nic
>

 

[Nero]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

No it's not................
like I said that reading is after onboard(MB)regulation.
I connected mine to a different MB and got a different reading.
Unload the 5v rail and then see.

 

[Paul]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

In article <qyUxc.13166$NK4.1896746@stones.force9.net>, "Nero"
<nero@rome.it> wrote:

> No it's not................
> like I said that reading is after onboard(MB)regulation.
> I connected mine to a different MB and got a different reading.
> Unload the 5v rail and then see.

What you are seeing, is not onboard regulation. The +5V from
the PSU is connected directly to the monitor chip. It is not
modified before it gets to that chip.

Onboard regulation is used to create new voltages, that don't
come from the power supply. Or to create voltages that can be
programmed by the BIOS or in response to a hardware signal on
the board. For example, +3.3V comes from the power supply,
but won't be used directly to power AGP I/O. That comes from
a separate regulated supply, but that supply needs the ability
to generate either 3.3V or 1.5V, depending on the TYPEDET# signal
on the AGP connector. So, that is an example of where a
programmable feature is needed, so the PSU power cannot be
used directly.

The reason you see the PSU voltages change, as you move from
motherboard to motherboard, has to do with the tolerances of
the voltage measurement done by the monitor chip. The monitor
chip on the motherboard is uncalibrated. If motherboard
manufacturers cared, they could feed precise voltages to the
circuit during manufacture, and work out compensating gain
and offset values, to improve the accuracy of the readings.
They don't do this, because there is no where to store the
info (a BIOS flash would wipe it out).

http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/627hf.pdf

For example, the Winbond W83627THF has an eight bit ADC for
voltage measurement, and a full scale value of 4.096V on
input. The minimum step size is 0.016V. The converter is
successive approximation type, so uses a DAC to make voltages
and does a binary search at 22KHz to try and match the
voltage it is measuring. The DAC is powered by an internal
reference called Vref, and that will have an accuracy, as to
how well it keeps its value. The DAC will use switched resistors,
and they have to be well matched, for the voltages it creates
to be smooth with respect to the eight bit number it is working
with. A typical chip Vref is only good to maybe 2%, or lower
tolerances are possible if devices are sorted according to
how well they make a reference voltage. Like all fine quality
datasheets, Winbond doesn't state how well the circuit works.

So, you have maybe 2.5% accuracy to start with. For voltages
outside the Winbond chip, requiring a voltage divider, to
make the voltage fit within the 4.096V measurement range,
two 1% resistors will be used. Their tolerances add, so now
the accuracy of the measurement is around 4.5% worst case.

Let us say you are interested in the voltage of +5V. On one
motherboard, the reading could be 4.5% low, and read 4.775V
On another motherboard, the reading could be 4.5% high, and
read 5.225V. Even though it is the same voltage from the
same PSU.

If you buy a multimeter, the multimeter is calibrated at the
factory, and that allows more accurate readings than the
motherboard circuit. The cheap meters I own are good for
2% plus 1 LSB or so. Multimeters have at least the equivalent
of a 12 bit ADC, which has a much better step size than the
8 bit SAR ADC of the Winbond chip. That is why measuring
the power supply with a meter is so important, as the
meter can read very accurately (the more money you have to
spend, the better they get).

"Homey", the guy who repairs motherboards, measures PSUs and
finds them to be doing much better than the motherboard
monitor would have you believe. I think he owns a good meter.

So, find a multimeter, and measure the supply while it is
under load. It is easy to get to the +5 and +12 voltages,
by using a drive power connector. The +3.3V can be read
from a 6 pin AUX power connector. The -5V and -12V don't
matter, and the reading from the monitor chip is good
enough, to at least determine they aren't grossly
malfunctioning.

Also, try unplugging the ATX20 pin connector, and then
replugging it (with PSU unplugged). Sometimes the contacts
aren't making good contact, and reseating the connector
will help for a while.

HTH,
Paul

 

[john]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

On Thu, 10 Jun 2004 06:10:14 -0400, nospam@needed.com (Paul) wrote:

>...
>So, find a multimeter, and measure the supply while it is
>under load.

Yeah, and don't forget to put in a fresh battery.

I hooked up a 9$ meter and read 16 volts on my 12 volt line.
Blinked twice, figured it out, and haven't worried about it since.

Gonna have to get a new battery in there someday

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

On Thu, 10 Jun 2004 06:10:14 -0400, nospam@needed.com (Paul) wrote:

>In article <qyUxc.13166$NK4.1896746@stones.force9.net>, "Nero"
><nero@rome.it> wrote:
>
>> No it's not................
>> like I said that reading is after onboard(MB)regulation.
>> I connected mine to a different MB and got a different reading.
>> Unload the 5v rail and then see.
>
>What you are seeing, is not onboard regulation. The +5V from
>the PSU is connected directly to the monitor chip. It is not
>modified before it gets to that chip.
>
>Onboard regulation is used to create new voltages, that don't
>come from the power supply. Or to create voltages that can be
>programmed by the BIOS or in response to a hardware signal on
>the board. For example, +3.3V comes from the power supply,
>but won't be used directly to power AGP I/O. That comes from
>a separate regulated supply, but that supply needs the ability
>to generate either 3.3V or 1.5V, depending on the TYPEDET# signal
>on the AGP connector. So, that is an example of where a
>programmable feature is needed, so the PSU power cannot be
>used directly.
>
>The reason you see the PSU voltages change, as you move from
>motherboard to motherboard, has to do with the tolerances of
>the voltage measurement done by the monitor chip. The monitor
>chip on the motherboard is uncalibrated. If motherboard
>manufacturers cared, they could feed precise voltages to the
>circuit during manufacture, and work out compensating gain
>and offset values, to improve the accuracy of the readings.
>They don't do this, because there is no where to store the
>info (a BIOS flash would wipe it out).
>
>http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/627hf.pdf
>
>For example, the Winbond W83627THF has an eight bit ADC for
>voltage measurement, and a full scale value of 4.096V on
>input. The minimum step size is 0.016V. The converter is
>successive approximation type, so uses a DAC to make voltages
>and does a binary search at 22KHz to try and match the
>voltage it is measuring. The DAC is powered by an internal
>reference called Vref, and that will have an accuracy, as to
>how well it keeps its value. The DAC will use switched resistors,
>and they have to be well matched, for the voltages it creates
>to be smooth with respect to the eight bit number it is working
>with. A typical chip Vref is only good to maybe 2%, or lower
>tolerances are possible if devices are sorted according to
>how well they make a reference voltage. Like all fine quality
>datasheets, Winbond doesn't state how well the circuit works.
>
>So, you have maybe 2.5% accuracy to start with. For voltages
>outside the Winbond chip, requiring a voltage divider, to
>make the voltage fit within the 4.096V measurement range,
>two 1% resistors will be used. Their tolerances add, so now
>the accuracy of the measurement is around 4.5% worst case.
>
>Let us say you are interested in the voltage of +5V. On one
>motherboard, the reading could be 4.5% low, and read 4.775V
>On another motherboard, the reading could be 4.5% high, and
>read 5.225V. Even though it is the same voltage from the
>same PSU.
>
>If you buy a multimeter, the multimeter is calibrated at the
>factory, and that allows more accurate readings than the
>motherboard circuit. The cheap meters I own are good for
>2% plus 1 LSB or so. Multimeters have at least the equivalent
>of a 12 bit ADC, which has a much better step size than the
>8 bit SAR ADC of the Winbond chip. That is why measuring
>the power supply with a meter is so important, as the
>meter can read very accurately (the more money you have to
>spend, the better they get).
>
>"Homey", the guy who repairs motherboards, measures PSUs and
>finds them to be doing much better than the motherboard
>monitor would have you believe. I think he owns a good meter.
>
>So, find a multimeter, and measure the supply while it is
>under load. It is easy to get to the +5 and +12 voltages,
>by using a drive power connector. The +3.3V can be read
>from a 6 pin AUX power connector. The -5V and -12V don't
>matter, and the reading from the monitor chip is good
>enough, to at least determine they aren't grossly
>malfunctioning.
>
>Also, try unplugging the ATX20 pin connector, and then
>replugging it (with PSU unplugged). Sometimes the contacts
>aren't making good contact, and reseating the connector
>will help for a while.
>
>HTH,
> Paul


lots of good info here thanks, atx connector was already reseated,
also bypassed the APC unit (just in case), removed my modem, my backup
120gig hdd, and my dvd burner. so... running 1 hdd, 1 cd/rw, 9700pro,
i'm still getting 4.48-4.57 on my +5v rail.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.asus (More info?)

> lots of good info here thanks, atx connector was already reseated,
> also bypassed the APC unit (just in case), removed my modem, my backup
> 120gig hdd, and my dvd burner. so... running 1 hdd, 1 cd/rw, 9700pro,
> i'm still getting 4.48-4.57 on my +5v rail.

Ask a friend for a new PSU to try or buy a new one: a system can't never
run stable with that low +5.

U need to check whats goin on.