Velocity Reviews

Velocity Reviews (http://www.velocityreviews.com/forums/index.php)
-   Computer Information (http://www.velocityreviews.com/forums/f41-computer-information.html)
-   -   Re: Power Supply Tester (http://www.velocityreviews.com/forums/t699029-re-power-supply-tester.html)

baron 09-22-2009 09:04 AM

Re: Power Supply Tester
 
James D. Andrews Inscribed thus:

> I should have asked here for recommendations before ordering, but I
> finally
> got around to ordering a cheap PSU tester last night. I'm interested
> in your experiences with these.
>
> The model I ordered was a Rexus PST-3 Power Supply Tester with LCD for
> $21. I read all of the 170+ reviews, and all were 4-5 stars except for
> less than a handful of bad experiences (a couple obviously from not
> following the
> directions). The biggest complaint was about the skimpy,
> foreign-tongue manual.
>
> I can't afford the $800 supermodel that tests everything all at once
> and makes coffee, so I had to go with an affordable model that like
> most
> products I'm sure will break down over many years of use. The old you
> get what you pay for thing here, but it was priced comparitively to
> others and it was not the cheapest model.
>
> So, do you think I'm going to regret this? Are these devices as
> useful as
> they say? Anyone have any major problems with this or other PSU
> testers? Anything I need to be particularly concerned about with using
> one of these?


Generally speaking the vast majority of these devices are little more
than a multimeter and a minimal load !

In order to properly test a computer PSU, you need to be able to load it
quite heavily. To at least to 50% of its maximum rating and then
measure the rail voltages. Not always easy to do properly.

I use 12volt car bulbs as loads and can easily get the 150 - 200 watts
or more that are needed to test properly. I've also found that many
PSU don't provide the rated total power output !

Your tester is fine for working/not working tests, but a multimeter
would tell you that very quickly !

--
Best Regards:
Baron.

Baron 09-22-2009 09:31 PM

Re: Power Supply Tester
 
James D. Andrews wrote:

> "baron" <baron.nospam@linuxmaniac.nospam.net> wrote in message
> news:h9a3uq$14f$1@news.eternal-september.org...
>> James D. Andrews Inscribed thus:


>>> The model I ordered was a Rexus PST-3 Power Supply Tester with LCD
>>> for $21.

>
>
> SNIP SNIP
>
>> Generally speaking the vast majority of these devices are little more
>> than a multimeter and a minimal load !
>>
>> In order to properly test a computer PSU, you need to be able to load
>> it quite heavily. To at least to 50% of its maximum rating and then
>> measure the rail voltages. Not always easy to do properly.
>>
>> I use 12volt car bulbs as loads and can easily get the 150 - 200
>> watts or more that are needed to test properly. I've also found that
>> many PSU don't provide the rated total power output !
>>
>> Your tester is fine for working/not working tests, but a multimeter
>> would tell you that very quickly !
>>
>> --
>> Best Regards:
>> Baron.

>
> Thanks. What I was going for was quick, easy, and almost idiot-proof
> (the biggest requirement considering the user - me).


<Grin> I've nothing against easy ! :-)

> Now, please explain to a simpleton like me, how do you have your car
> lights set up.


Simply connect as many bulbs as you need in parallel to get the number
of watts you want. Remember that you have 12v 5v and 3.3volt rails.
The rating plate on the PSU should give you either the total power or
the number of amps per voltage rail.

The 12volt rail is easy ! Simply look at the bulb wattage. For the 5
and 3.3 volt rails a little math is needed to calculate the Wattage and
current. Although a quick guide would be to divide the wattage by two
for the 5volt rail and three for the 3.3 volt one.

Ohms law gives:- I = P/E and R = E/I also R = P/I^2
Where I is Current, P is Power, R is Resistance and E is voltage.

So a 60watt bulb at 12volts (60/12)=5amps. A resistance of 2.4ohms.
The same bulb at 5volts ((5^2)/2.4)=10.4watts and draw 2.08amps.
The same bulb at 3.3 volts ((3.3^2)/2.4)=4.5watts and draw 1.37amps.

Once you know the parameters for any particular bulb you can use almost
anything. I've even used electric fire bar elements to make power
loads.

HTH.

--
Best Regards:
Baron.

baron 09-23-2009 10:19 AM

Re: Power Supply Tester
 
Robert Baer Inscribed thus:

> baron wrote:
>> James D. Andrews Inscribed thus:
>>
>>> I should have asked here for recommendations before ordering, but I
>>> finally
>>> got around to ordering a cheap PSU tester last night. I'm
>>> interested in your experiences with these.
>>>
>>> The model I ordered was a Rexus PST-3 Power Supply Tester with LCD
>>> for $21. I read all of the 170+ reviews, and all were 4-5 stars
>>> except for less than a handful of bad experiences (a couple
>>> obviously from not following the
>>> directions). The biggest complaint was about the skimpy,
>>> foreign-tongue manual.
>>>
>>> I can't afford the $800 supermodel that tests everything all at once
>>> and makes coffee, so I had to go with an affordable model that like
>>> most
>>> products I'm sure will break down over many years of use. The old
>>> you get what you pay for thing here, but it was priced comparitively
>>> to others and it was not the cheapest model.
>>>
>>> So, do you think I'm going to regret this? Are these devices as
>>> useful as
>>> they say? Anyone have any major problems with this or other PSU
>>> testers? Anything I need to be particularly concerned about with
>>> using one of these?

>>
>> Generally speaking the vast majority of these devices are little more
>> than a multimeter and a minimal load !
>>
>> In order to properly test a computer PSU, you need to be able to load
>> it quite heavily. To at least to 50% of its maximum rating and then
>> measure the rail voltages. Not always easy to do properly.
>>
>> I use 12volt car bulbs as loads and can easily get the 150 - 200
>> watts or more that are needed to test properly. I've also found that
>> many PSU don't provide the rated total power output !
>>
>> Your tester is fine for working/not working tests, but a multimeter
>> would tell you that very quickly !
>>

> Technically speaking, DC testing is incomplete..supply voltages
> could be in spec at load, but be too noisy (ie: out of spec).
> But that is rare..


I must admit I've had PSU that whilst operating quite normally, produced
enough RF hash to drown out radio and in one case interfere with TV
reception.

--
Best Regards:
Baron.

Paul 09-23-2009 01:11 PM

Re: Power Supply Tester
 
baron wrote:
> Robert Baer Inscribed thus:
>
>> baron wrote:
>>> James D. Andrews Inscribed thus:
>>>
>>>> I should have asked here for recommendations before ordering, but I
>>>> finally
>>>> got around to ordering a cheap PSU tester last night. I'm
>>>> interested in your experiences with these.
>>>>
>>>> The model I ordered was a Rexus PST-3 Power Supply Tester with LCD
>>>> for $21. I read all of the 170+ reviews, and all were 4-5 stars
>>>> except for less than a handful of bad experiences (a couple
>>>> obviously from not following the
>>>> directions). The biggest complaint was about the skimpy,
>>>> foreign-tongue manual.
>>>>
>>>> I can't afford the $800 supermodel that tests everything all at once
>>>> and makes coffee, so I had to go with an affordable model that like
>>>> most
>>>> products I'm sure will break down over many years of use. The old
>>>> you get what you pay for thing here, but it was priced comparitively
>>>> to others and it was not the cheapest model.
>>>>
>>>> So, do you think I'm going to regret this? Are these devices as
>>>> useful as
>>>> they say? Anyone have any major problems with this or other PSU
>>>> testers? Anything I need to be particularly concerned about with
>>>> using one of these?
>>> Generally speaking the vast majority of these devices are little more
>>> than a multimeter and a minimal load !
>>>
>>> In order to properly test a computer PSU, you need to be able to load
>>> it quite heavily. To at least to 50% of its maximum rating and then
>>> measure the rail voltages. Not always easy to do properly.
>>>
>>> I use 12volt car bulbs as loads and can easily get the 150 - 200
>>> watts or more that are needed to test properly. I've also found that
>>> many PSU don't provide the rated total power output !
>>>
>>> Your tester is fine for working/not working tests, but a multimeter
>>> would tell you that very quickly !
>>>

>> Technically speaking, DC testing is incomplete..supply voltages
>> could be in spec at load, but be too noisy (ie: out of spec).
>> But that is rare..

>
> I must admit I've had PSU that whilst operating quite normally, produced
> enough RF hash to drown out radio and in one case interfere with TV
> reception.
>


You can buy common mode filters that would suppress the noise back on AC,
but they aren't cheap. They come in a nice metal box though.

http://media.digikey.com/PDF/Data%20...%20Filters.pdf

For example, the good one in that datasheet, is $66.

http://search.digikey.com/scripts/Dk...408328&k=10vn1

What Robert might have been referring to, is ripple on the DC output side.

In this power supply review, you can see some pictures at the bottom of the
page, that show ripple riding on the DC outputs, under various conditions.

http://www.jonnyguru.com/modules.php...Story2&reid=24

So a power supply can send noise two ways. It can put ripple on the DC outputs.
But it can also drive noise back down the AC line, and the AC power cord becomes
an antenna. I know I have at least one power supply here, where the filtering
going back towards AC is poor. Sometimes, you can find a power strip that has
AC filtering, and that will be enough to attenuate the switching noise coming
back from the PSU.

Paul

baron 09-23-2009 03:02 PM

Re: Power Supply Tester
 
Paul Inscribed thus:

> baron wrote:
>> Robert Baer Inscribed thus:
>>
>>> baron wrote:
>>>> James D. Andrews Inscribed thus:
>>>>
>>>>> I should have asked here for recommendations before ordering, but
>>>>> I finally
>>>>> got around to ordering a cheap PSU tester last night. I'm
>>>>> interested in your experiences with these.
>>>>>
>>>>> The model I ordered was a Rexus PST-3 Power Supply Tester with LCD
>>>>> for $21. I read all of the 170+ reviews, and all were 4-5 stars
>>>>> except for less than a handful of bad experiences (a couple
>>>>> obviously from not following the
>>>>> directions). The biggest complaint was about the skimpy,
>>>>> foreign-tongue manual.
>>>>>
>>>>> I can't afford the $800 supermodel that tests everything all at
>>>>> once and makes coffee, so I had to go with an affordable model
>>>>> that like most
>>>>> products I'm sure will break down over many years of use. The old
>>>>> you get what you pay for thing here, but it was priced
>>>>> comparitively to others and it was not the cheapest model.
>>>>>
>>>>> So, do you think I'm going to regret this? Are these devices as
>>>>> useful as
>>>>> they say? Anyone have any major problems with this or other PSU
>>>>> testers? Anything I need to be particularly concerned about with
>>>>> using one of these?
>>>> Generally speaking the vast majority of these devices are little
>>>> more than a multimeter and a minimal load !
>>>>
>>>> In order to properly test a computer PSU, you need to be able to
>>>> load
>>>> it quite heavily. To at least to 50% of its maximum rating and
>>>> then
>>>> measure the rail voltages. Not always easy to do properly.
>>>>
>>>> I use 12volt car bulbs as loads and can easily get the 150 - 200
>>>> watts or more that are needed to test properly. I've also found
>>>> that many PSU don't provide the rated total power output !
>>>>
>>>> Your tester is fine for working/not working tests, but a multimeter
>>>> would tell you that very quickly !
>>>>
>>> Technically speaking, DC testing is incomplete..supply voltages
>>> could be in spec at load, but be too noisy (ie: out of spec).
>>> But that is rare..

>>
>> I must admit I've had PSU that whilst operating quite normally,
>> produced enough RF hash to drown out radio and in one case interfere
>> with TV reception.
>>

>
> You can buy common mode filters that would suppress the noise back on
> AC, but they aren't cheap. They come in a nice metal box though.
>
>

http://media.digikey.com/PDF/Data%20...%20Filters.pdf
>
> For example, the good one in that datasheet, is $66.
>
>

http://search.digikey.com/scripts/Dk...408328&k=10vn1
>
> What Robert might have been referring to, is ripple on the DC output
> side.
>
> In this power supply review, you can see some pictures at the bottom
> of the page, that show ripple riding on the DC outputs, under various
> conditions.
>
> http://www.jonnyguru.com/modules.php...Story2&reid=24
>
> So a power supply can send noise two ways. It can put ripple on the DC
> outputs. But it can also drive noise back down the AC line, and the AC
> power cord becomes an antenna. I know I have at least one power supply
> here, where the filtering going back towards AC is poor. Sometimes,
> you can find a power strip that has AC filtering, and that will be
> enough to attenuate the switching noise coming back from the PSU.
>
> Paul


Thanks for the links. Quite interesting.
As far as the noisy (RF Radiating) PSU are concerned, I just bang them
back at the supplier as faulty. Which arguably they are.

--
Best Regards:
Baron.

baron 09-23-2009 03:15 PM

Re: Power Supply Tester
 
Paul Inscribed thus:

>
> In this power supply review, you can see some pictures at the bottom
> of the page, that show ripple riding on the DC outputs, under various
> conditions.
>
> http://www.jonnyguru.com/modules.php...Story2&reid=24
>
>
> Paul


Just a comment on this particular link. I had to study a little but I
found it odd that the noise for the 12volt outputs varied for each
12volt supply in each test. Internally all the 12volt wires are
connected to the same place.

--
Best Regards:
Baron.

Paul 09-23-2009 07:15 PM

Re: Power Supply Tester
 
baron wrote:
> Paul Inscribed thus:
>
>> In this power supply review, you can see some pictures at the bottom
>> of the page, that show ripple riding on the DC outputs, under various
>> conditions.
>>
>> http://www.jonnyguru.com/modules.php...Story2&reid=24
>>
>>
>> Paul

>
> Just a comment on this particular link. I had to study a little but I
> found it odd that the noise for the 12volt outputs varied for each
> 12volt supply in each test. Internally all the 12volt wires are
> connected to the same place.
>


There are other questions you can ask as well. The ripple signal
is at 2.5KHz, and what relationship does that have to power
supply operation ? The switching must run at a higher frequency than
that.

There are parameters recorded on the scope trace screenshot.
The scope is sampling at 25K samples/sec, the time base is
2 milliseconds per division. There would be 50 samples per
division at 40 microseconds each. There are five ripple sine
waves per division. There are 10 digital samples per waveform.
This is higher than Nyquist, and would be my bare minimum sampling
rate for a decent waveform. (10x sample rate, 2.5KHz noise, 25KSa/sec
sampling rate).

A couple of the traces, seem to have some other noise frequency, higher
than 2.5KHz.

One thing you have to be careful of, when making a ripple measurement,
is that the scope leads don't pick up any stray noise. The ATX spec
even tells you how to do that (page 16 differential noise).

http://www.formfactors.org/developer...X12V_1_3dg.pdf

They're using a P6046 probe to do the measurement in the ATX spec, and
when I look that up, it is a diff probe. (I think I've even used
one of those on occasion.) Using a differential probe, is intended to
remove common mode noise, such as noise coupled into the measurement
leads. Is Jonny's rig that good, or is it an "A-B" measurement, using
two single ended probes, or even a single ended measurement which could
pick up noise along the way ?

http://www.tucker.com/java/jsp/produ...0281_condR.htm

But in terms of giving a rough amplitude, the measurements are good enough :-)
I wouldn't waste my time complaining about them :-) Lord knows I've had
enough trouble doing measurements like that in the lab myself, and
Jonny's results look pretty clean compared to some messes I've had.

Paul

baron 09-24-2009 11:30 AM

Re: Power Supply Tester
 
James D. Andrews Inscribed thus:

>
> "Baron" <baron.nospam@linuxmaniac.nospam.net> wrote in message
> news:h9bfn8$oav$1@news.eternal-september.org...
>> James D. Andrews wrote:
>>
>>> "baron" <baron.nospam@linuxmaniac.nospam.net> wrote in message
>>> news:h9a3uq$14f$1@news.eternal-september.org...
>>>> James D. Andrews Inscribed thus:

>>
>>>>> The model I ordered was a Rexus PST-3 Power Supply Tester with LCD
>>>>> for $21.
>>>
>>>
>>> SNIP SNIP
>>>
>>>> Generally speaking the vast majority of these devices are little
>>>> more than a multimeter and a minimal load !
>>>>
>>>> In order to properly test a computer PSU, you need to be able to
>>>> load
>>>> it quite heavily. To at least to 50% of its maximum rating and
>>>> then
>>>> measure the rail voltages. Not always easy to do properly.
>>>>
>>>> I use 12volt car bulbs as loads and can easily get the 150 - 200
>>>> watts or more that are needed to test properly. I've also found
>>>> that many PSU don't provide the rated total power output !
>>>>
>>>> Your tester is fine for working/not working tests, but a multimeter
>>>> would tell you that very quickly !
>>>>
>>>> --
>>>> Best Regards:
>>>> Baron.
>>>
>>> Thanks. What I was going for was quick, easy, and almost
>>> idiot-proof (the biggest requirement considering the user - me).

>>
>> <Grin> I've nothing against easy ! :-)
>>
>>> Now, please explain to a simpleton like me, how do you have your car
>>> lights set up.

>>
>> Simply connect as many bulbs as you need in parallel to get the
>> number
>> of watts you want. Remember that you have 12v 5v and 3.3volt rails.
>> The rating plate on the PSU should give you either the total power or
>> the number of amps per voltage rail.
>>
>> The 12volt rail is easy ! Simply look at the bulb wattage. For the 5
>> and 3.3 volt rails a little math is needed to calculate the Wattage
>> and
>> current. Although a quick guide would be to divide the wattage by
>> two for the 5volt rail and three for the 3.3 volt one.
>>
>> Ohms law gives:- I = P/E and R = E/I also R = P/I^2
>> Where I is Current, P is Power, R is Resistance and E is voltage.
>>
>> So a 60watt bulb at 12volts (60/12)=5amps. A resistance of 2.4ohms.
>> The same bulb at 5volts ((5^2)/2.4)=10.4watts and draw 2.08amps.
>> The same bulb at 3.3 volts ((3.3^2)/2.4)=4.5watts and draw 1.37amps.
>>
>> Once you know the parameters for any particular bulb you can use
>> almost
>> anything. I've even used electric fire bar elements to make power
>> loads.
>>
>> HTH.
>>
>> --
>> Best Regards:
>> Baron.

>
>
> Many thanks


You're welcome.

--
Best Regards:
Baron.


All times are GMT. The time now is 07:29 PM.

Powered by vBulletin®. Copyright ©2000 - 2014, vBulletin Solutions, Inc.
SEO by vBSEO ©2010, Crawlability, Inc.