The time to discharge method is good.
You can also do an in circuit test.
Measure the ripple Vpp using an oscilloscope.
Alternatively, but this is not as reliable, use the AC scale of your DMM to measure the average ripple on the DC voltage. Then triple that to get very roughly the Vpp on the supply rail.
If you have one good rail and one bad rail, the difference will show up. But this in circuit measurement will also show differences due to different current draw on the supply rail.
If you build an esr meter, you can do an in circuit esr measurement.
Google diy esr meter.
You can also do an in circuit test.
Measure the ripple Vpp using an oscilloscope.
Alternatively, but this is not as reliable, use the AC scale of your DMM to measure the average ripple on the DC voltage. Then triple that to get very roughly the Vpp on the supply rail.
If you have one good rail and one bad rail, the difference will show up. But this in circuit measurement will also show differences due to different current draw on the supply rail.
If you build an esr meter, you can do an in circuit esr measurement.
Google diy esr meter.
Last edited:
How does the discharge method differentiate between a cap with normal ESR and one with high ESR ?
Unless I'm missing something, this method seems a little like the equivalent of testing batteries by measuring the open circuit voltage. The internal resistance isn't accounted for.
Many faulty electrolytics still have their nominal capacity, its just that the ESR has risen dramatically.
Unless I'm missing something, this method seems a little like the equivalent of testing batteries by measuring the open circuit voltage. The internal resistance isn't accounted for.
Many faulty electrolytics still have their nominal capacity, its just that the ESR has risen dramatically.
There is every chance that the capacity will be reduced if the ESR is high. It is meant only as a guide for first aid. Using a capacitance meter is the choice of all professionals but to the amateur, with no other method, the system works well enough to indicate a problem.
Hmmm. I remain to be convinced on that. I never found a cap meter much use at all when faultfinding tbh (and it used to be the day job
). Maybe some of this is meter specific but I do remember adding series resistance to caps in an attempt to see how it altered readings. You could add crazy amounts (10's to 100's of ohms) before any real change in capacity showed.
If capacitor is of reduced capacity because holes have formed in the foils, then that is also likely to have increased esr.
I don't know which test (esr or capacity) is more sensitive to a gradually wearing out electrolytic.
Would a 20% reduction in capacity show more performance deterioration than a 20% increase in esr?
I don't know that either.
Maybe we need to have a list of tests and find which give an earlier, or better, indication of a failing electrolytic.
I don't know which test (esr or capacity) is more sensitive to a gradually wearing out electrolytic.
Would a 20% reduction in capacity show more performance deterioration than a 20% increase in esr?
I don't know that either.
Maybe we need to have a list of tests and find which give an earlier, or better, indication of a failing electrolytic.
Is this is a question for me?
Yes,this was a 12v filter capacitor on the secondary side on a smps and first test with a scope shows clearly a big ripple. Then the capacitor removed discharged and measured with different equipments as above.
I would say in the majority of cases, no, a drop in capacity would have minimal effect.
I used to encounter this on a daily basis. A quick confirmation of a suspect cap would be just to slap another across the suspect device. Even something like a 10uf for 100uf or a 220uf for 1000uf proved the point conclusively that the suspect cap was duff. Checking the removed part on a cap meter often showed the faulty device was OK value wise.
Don't I need to take "the suspect" out of the circuit first? The other problem is that I don't know which of the two caps is "the suspect." I do have a parts unit with a functioning power supply, so maybe I'll just swap in a good cap for each of the two suspects and see if that works. BTW the buzz is so loud that it clips both channels. Am I on the right track or could it be the rectifier bridge that went out?
Last edited:
"Don't I need to take "the suspect" out of the circuit first? The other problem is that I don't know which of the two caps is "the suspect." "
A cap can fail (1) effectively open-circuit, (2) effectively short-circuit or (3) something roughly in between but with a very much changed value.
Case (1), added parallel cap will cure problem completely.
Case (2), there would have been a bang, smoke, blown fuse then silence.
Case (3), added cap should restore working correctly.
If adding cap makes no difference, move on to another one!
This is so much easier with diagnostic tools like an oscilloscope, ESR meter etc
A cap can fail (1) effectively open-circuit, (2) effectively short-circuit or (3) something roughly in between but with a very much changed value.
Case (1), added parallel cap will cure problem completely.
Case (2), there would have been a bang, smoke, blown fuse then silence.
Case (3), added cap should restore working correctly.
If adding cap makes no difference, move on to another one!
This is so much easier with diagnostic tools like an oscilloscope, ESR meter etc
Thanks. That's down to to many years spent as a bench tech
It can stay in place because the failure mode is simply that it has become less effective. Its not pulling anything down or suspected of drawing excess current (that would cause it to get hot and go pop). The only caution I would add is that with large caps you should switch off first and then tag the new cap across the suspect rather than really trying to dab it across handheld fashion
Just do one at a time. If one of the diodes in bridge was short or leaky then the bridge would be getting extremely hot and probably failing totally within a few seconds. If a diode was open then ripple would increase dramatically under load. It could happen but a bridge failing in this way would be unusual.
I'm going to give this a try, but aren't case 1 and case 3 pretty much the same?
BGW 620B Hum?
Hi audiomagnate
1. I designed this Amplifier and many others while @ BGW. Looking @ your questions you may not have strong tech training or equipment. So let’s try some steps.
2. Do you have the manual?
3. Have you modified the 620B for Mono mode or is it still Stereo?
4. If Mono mode then the problem could be the driven channel (Left). Remove the Octal connector to the Right module. Turn on the amplifier and see if the “HUM” is present.
5. If the amplifier is stock the PS cap or a broken wire maybe the problem.
6. Make sure that no connections other than AC power and a speaker are connected. You may want to add a series resistor to the speaker to test to reduce any large power hum to them during any testing.
7. Can you measure the ± voltages on the Big caps? Are they the same? If they are the same, the caps are still good.
Duke
Hi audiomagnate
1. I designed this Amplifier and many others while @ BGW. Looking @ your questions you may not have strong tech training or equipment. So let’s try some steps.
2. Do you have the manual?
3. Have you modified the 620B for Mono mode or is it still Stereo?
4. If Mono mode then the problem could be the driven channel (Left). Remove the Octal connector to the Right module. Turn on the amplifier and see if the “HUM” is present.
5. If the amplifier is stock the PS cap or a broken wire maybe the problem.
6. Make sure that no connections other than AC power and a speaker are connected. You may want to add a series resistor to the speaker to test to reduce any large power hum to them during any testing.
7. Can you measure the ± voltages on the Big caps? Are they the same? If they are the same, the caps are still good.
Duke
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.