how to evaluate the data variation before and after environment test

[hhorse]

when we do environmental stress test for PSU products, we'll perform function check before and after test. The function check include regulation, ripple current measurements etc..
Generally, the parametric data will change between 2 function checks even though the data are all within spec requirement. Just want to know if there's any acception rules on the data difference before/after environment test?
for example, PSU output voltage range=(11.4~12.6V)
case 1: PSU output voltage before EST=11.85V
PSU output after EST=12.05V
Delta%=1.7%
case 2 PSU output voltage before EST=11.85V
PSU output after EST=12.48V
Delta%=5.3%
can we say something wrong is with case 2? or what dose the big delta imply?

 

[IRstuff]

Where's the documentation that the test conditions are the same? If you are a the ESS stage, you should already have the performance verification against temperature, and you should be able to determine whether the change is consistent with your test conditions. Frankly, even your 1.7% variation is rather large for what should have been an infant mortality and workmanship test.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[zappedagain]

How long after the EST did you wait before you tested? If any components are still 'hot or cold' or 'soggy or dry' that can throw off your test. It really depends on what you are trying to test.

Z

 

[hhorse]

thanks IRstuff & zappedagain.
looks you both think the variation should be very small at the same conditions..
my concern is how to define the "degree". how much small value will be acceptable? otherwise it shows the PSU is abnormal.

 

[IRstuff]

That's a question for your designers, isn't it? Alternately, you could run 100 power cycles and measure the variance.

But, fundamentally, ESS is typically a very short and non-life-consuming test, so if such a brief test evokes even a 1.7% change, what's stopping it from going to 10% change a year from now? What's the end-of-life spec? Is your post-ESS spec toleranced for end-of-life and component variances over time? If not, you ought to have a toleranced spec that allows for that. In your case, I would think that anything over 1% is troubling, given what appears to be a 12V ± 0.6V spec. Certainly, your second unit at 12.48V is so close to the upper limit that you have to question whether it won't go out of spec before its warranty period, which is a cost adder.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529