Power Supply Voltage Tolerances and Measurements

Electronics Computer Programming Q&A
User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: Power Supply Voltage Tolerances and Measuremen

Post by MrAl » Wed Oct 28, 2009 4:53 am

Hello again,


I just had to add this note too...

When i look at some power supplies cheap or expensive, i see MTBF (mean time
between failures) spec'd at 100,000 hours. Yes, that is one hundred thousand
hours, not 10,000 hours. What i want to know is, where are they finding
the caps to do this with ha ha.
I guess they figure it wont have to operate at the full temperature of 105
degrees so it will last much longer.
LEDs vs Bulbs, LEDs are winning.

User avatar
reloadron
Posts: 519
Joined: Sat Jun 28, 2008 8:57 am
Location: Cleveland, Ohio
Contact:

Re: Power Supply Voltage Tolerances and Measurements

Post by reloadron » Wed Oct 28, 2009 4:33 pm

Seems that they sort of conveniently forget to tell the end user or buyer a few things. When I was helping with PSU testing something I noticed of interest was that the output power ratings that many manufacturers bragged about were slightly inflated. For example the labeled power out was based on a room temperature (PSU Inlet Temp) of about 23 degrees C. or about 73.4 degrees F.

The problem is that just about every home computer PSU I have ever seen does not enjoy running at those temperatures. The PSU in most home computers draws its cooling air (inlet) from inside the case. With the CPU, Graphics Card, and Chipset the temperatures inside the case are much warmer, especially on larger complex systems. I did a little testing along those lines and measured some temperatures. I looked at case inlet (room ambient), case outlet (rear exhaust), PSU Inlet (at the PSU inlet fan) and PSU Outlet (at the PSU exhaust).

Typical PSU inlet temperatures were around 95 to 100 degrees F. That was with a beefy machine running 100% CPU usage. Room temperature was around 75 degrees F. when I did the experiments. I used 4 thermocouples to measure the temperatures.

So how much should the PSU rated power be derated for those inlet temperatures? The project was done for a computer enthusiast forum. The remains of the pictures and a few screen shots can be found here. The very bottom screen shot were the temperatures with the processors running at 100% and done after about 20 min to allow the internal temperatures to get stable.

Bottom line is short of sitting on a bench no PSU is cooled with room ambient air. Interesting that the PSU overall added little heat based on PSU Inlet and PSU Outlet.

Ron

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: Power Supply Voltage Tolerances and Measuremen

Post by MrAl » Wed Oct 28, 2009 4:54 pm

Hi Ron,


Did they determine any derating factor for the power supplies then?

I guess 100 degrees F isnt that hot electronically speaking, being only
about 37.8 degrees C.
LEDs vs Bulbs, LEDs are winning.

User avatar
reloadron
Posts: 519
Joined: Sat Jun 28, 2008 8:57 am
Location: Cleveland, Ohio
Contact:

Re: Power Supply Voltage Tolerances and Measurements

Post by reloadron » Wed Oct 28, 2009 5:30 pm

Nope. On good commercial power supplies it seems finding derating data is easy but on those computer power supplies we were hard pressed to find useful data. If I recall correctly a few manufacturers do provide it but very few. Been awhile since that fad went through my life. :smile:

User avatar
CeaSaR
Posts: 1702
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Power Supply Voltage Tolerances and Measurements

Post by CeaSaR » Mon Mar 08, 2010 1:48 pm

Hey guys,

My HP was acting slow so I shut it down and removed power for a really hard reset. Took a few minutes to vacuum
the case etc., and went to turn it back on - nothing. Pulled the case apart, further cleaning, still nothing. Pulled the
PSU, cleaned and tried again, nothing but the blinky light on the back of the PSU. Further inspection showed all the
2200 uF 10 V 'lytics bulged. I'm off to scour the local RS's for caps. Need 6 in all. Thanks for the thread. I'll let you
know how it went later.

CeaSaR
Hey, what do I know?

Robert Reed
Posts: 2276
Joined: Wed Nov 24, 2004 1:01 am
Location: ASHTABULA,OHIO
Contact:

Re: Power Supply Voltage Tolerances and Measuremen

Post by Robert Reed » Mon Mar 08, 2010 5:17 pm

Caesar
How old is your computer? How many days running time do you think were on it? It will be interesting to hear about your results.

User avatar
CeaSaR
Posts: 1702
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Power Supply Voltage Tolerances and Measurements

Post by CeaSaR » Mon Mar 08, 2010 7:59 pm

Ok, RS fails once again. Only (3) 2200 uF, but 35 V and axial leads. :x I had little hope anyway. Time for mailorder.
Which leads to a question. I see a cap on the mobo that looks bad also. I think it is a 820 uF at 6.3 V. Does that
sound like a common value for a mobo? It is between the first 2 expansion slots, Video and PCI, I beleive.

Robert,
The HP is approximately 7 years old, and I'd say it has been on roughly 50% of it's life. That's a little over 30,000 hours.
Not to shabby, I guess. I think it should be longer though.

CeaSaR
Hey, what do I know?

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: Power Supply Voltage Tolerances and Measuremen

Post by MrAl » Tue Mar 09, 2010 4:29 am

Hi Ceasar,


The only caution i can think of here is that buying electrolytics from RS may not be a good idea.
The reason being that RS does not usually have spec's for their cap's ESR. The ESR is good to
know so that you can keep it low. When you buy from a place like Digikey you can look at the
data sheet to find out the ESR and it's usually spec'd as somewhat low when you have the right
caps selected. That's not to say RS caps wont work, but that it's a guessing game as to what
the ESR really is and so how much ripple will result at the output of the power supply.
This also means that any two caps of the same value may vary widely as to what their ESR
really is.
On the plus side, i had purchased a set of electrolytic caps from RS a number of years ago and
i used a few in a switching power supply and the output ripple didnt turn out too bad. I guess
i got lucky there as the ESR must have been low enough in at least the output cap. That was
for a battery charger though, not a computer power supply. Not sure if i would trust an
unknown ESR value for my computer without some really good tests.

As to the values themselves in a power supply, i have found that my power supply had lots of
different values and sizes of caps with different voltage ratings. The value of the cap (in uf)
was always stamped on the side, as well as the voltage rating (in volts). It was hard to see
some of the values though because other parts were in the way, but some perseverance paid
off. A small mirror helped in some cases.
LEDs vs Bulbs, LEDs are winning.

User avatar
CeaSaR
Posts: 1702
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Power Supply Voltage Tolerances and Measurements

Post by CeaSaR » Tue Mar 09, 2010 9:16 am

MrAl,

I didn't think RS would have any anyway, but I was willing to give them a try. I'm going to go to Mouser/Digikey/MCM
etal and look at the ones I am in need of. All the PSU caps I need are 2200uF 10V. The one on the motherboard looks
like it says 820uF 6.3V. The darn thing is right in between the Video slot and the neighboring PCI slot. It's very hard
to read, but that's what I could see. It has a "K" shaped "vent" on top. Does the mobo cap sound right?

Thanks for the help/advice.

CeaSaR
Hey, what do I know?

Robert Reed
Posts: 2276
Joined: Wed Nov 24, 2004 1:01 am
Location: ASHTABULA,OHIO
Contact:

Re: Power Supply Voltage Tolerances and Measuremen

Post by Robert Reed » Tue Mar 09, 2010 10:02 am

Caesar
"The one on the motherboard looks
like it says 820uF "
If you were wondering if that was a legit value-YES. It follows the EIA value code-i.e.----62,68,75,82.91.100-----.

User avatar
CeaSaR
Posts: 1702
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Power Supply Voltage Tolerances and Measurements

Post by CeaSaR » Tue Mar 09, 2010 10:53 am

I guess I'll have to pull the MOBO again and see if I can verify that value further. It's a dinky little thing though. I might
be able to clean it up and take a picture and blow it up (the picture, not the cap!). I'll let you know the result.

I'm looking at caps now. So far:
Mouser
Jameco
MCM
Digikey has me stumped at the filter page. At the values I need (2000uF/10V), the ESR's available are "-, 130mOhm, 140mOhm".
The series run from EB to Z, and from what I have been able to discern, I am interested in the K or KM series. At this point all
that is listed is KMG, KT, KW, KY and KZE. The closest is KMG, right? Are any of the others close? Although when I cick "In Stock"
it reduces the list down to 19 (from 53).

All help very greatly appreciated,

CeaSaR
Hey, what do I know?

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: Power Supply Voltage Tolerances and Measuremen

Post by MrAl » Tue Mar 09, 2010 12:02 pm

Hi again,


I found these on Digikey:
PCE4872CT-ND

they are a bit expensive (3 dollars each) though (look like the solid type).

There are cheaper ones:
P10230-ND

But you can also use 16v units if the size will still fit.

You could also check more into the solid type which are supposed to last longer.
LEDs vs Bulbs, LEDs are winning.

User avatar
CeaSaR
Posts: 1702
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Power Supply Voltage Tolerances and Measuremen

Post by CeaSaR » Tue Mar 09, 2010 12:55 pm

MrAl,

The PCE4872CT-ND says it's a SMD part, I need TH, which the P10230-ND is. Hmm, datasheet doesn't list ESR.
How about this one? It's at least in the K series, if that makes a difference.

Thanks again,

CeaSaR
Hey, what do I know?

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Re: Power Supply Voltage Tolerances and Measuremen

Post by MrAl » Tue Mar 09, 2010 2:23 pm

Hi again,


That doesnt look too bad i guess, not the best though, ESR around 0.16 ohms.
LEDs vs Bulbs, LEDs are winning.

User avatar
CeaSaR
Posts: 1702
Joined: Sat Nov 08, 2003 1:01 am
Location: Phoenixville, PA USA
Contact:

Re: Power Supply Voltage Tolerances and Measurements

Post by CeaSaR » Tue Mar 09, 2010 2:52 pm

Where'd you find the ESR? I looked on the datasheet and couldn't find it. It wasn't listed on the site either.
If the TH part you listed is OK, I'll go with that.

CeaSaR
Hey, what do I know?

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests