UK Vintage Radio Repair and Restoration Discussion Forum - View Single Post

mhennessy · 8th Mar 2016, 1:14 pm

Curious...

If pin 6 was low, that would be most unusual because that's directly from the bandgap generator, and these are usually very accurate because of fundamental laws of physics...

Perhaps leakage on the PCB would be a factor, but then your separate test would have eliminated that...

If both were out by the same amount, then your DMM would be the obvious suspect - the Fluke 73 does have an internal preset, so that is possible. However, your errors were 3% and 6% - as these were (presumably) measured with the same meter, it's unlikely to be the culprit. However, there is a chance that if your meter was out by 3%, then one IC was correct and the other was out by 3%. But that does seem unlikely, and your next step - whatever that might be - will help answer this one.

The fact the line regulation is OK suggests that at least some parts of the ICs were OK.

Were these the type that were tested before dispatch? If so, I presume the recorded figures were plausible?

Just about the only thing I'm left with is static discharge. But it would be extremely unlucky to damage two ICs - mounted on a PCB - in the same way. But what's left?

Oh - did you put a 'scope across the output to ensure it wasn't oscillating or picking up RF?

Would you like me to send you an LM4040? I have a fair few of them here. I'll measure the output first, and that will allow you to "reality check" your DVMs. FOC, naturally...

8th Mar 2016, 1:14 pm	#71
mhennessy Dekatron Join Date: Jul 2007 Location: Evesham, Worcestershire, UK. Posts: 4,244	Re: A precision voltage calibrator Curious... If pin 6 was low, that would be most unusual because that's directly from the bandgap generator, and these are usually very accurate because of fundamental laws of physics... Perhaps leakage on the PCB would be a factor, but then your separate test would have eliminated that... If both were out by the same amount, then your DMM would be the obvious suspect - the Fluke 73 does have an internal preset, so that is possible. However, your errors were 3% and 6% - as these were (presumably) measured with the same meter, it's unlikely to be the culprit. However, there is a chance that if your meter was out by 3%, then one IC was correct and the other was out by 3%. But that does seem unlikely, and your next step - whatever that might be - will help answer this one. The fact the line regulation is OK suggests that at least some parts of the ICs were OK. Were these the type that were tested before dispatch? If so, I presume the recorded figures were plausible? Just about the only thing I'm left with is static discharge. But it would be extremely unlucky to damage two ICs - mounted on a PCB - in the same way. But what's left? Oh - did you put a 'scope across the output to ensure it wasn't oscillating or picking up RF? Would you like me to send you an LM4040? I have a fair few of them here. I'll measure the output first, and that will allow you to "reality check" your DVMs. FOC, naturally...