Lately I have come across a desire to be able to test a capacitor, and can't do it with the one meter I have.
The capacitor in question is a round-can 440v 40/5 µF castor-oil-filled capacitor from an outside condenser of a central-AC unit. The indoor AC unit blows air, but it isn't cold.... Casual observation showed that the condenser unit fan will not start on its own, but it spins freely and will run if I spin the fan with a stick--and then the AC produces cold air like normal. The outdoor condenser unit was replaced only ~4 years ago, so I am guessing the cap just failed.
After shutting off the power to the AC system, I removed the capacitor and will just go shopping for a new one tomorrow. I got curious tho and decided to try testing the capacitor, since I had one DMM that could do that (I thought).
The best meter I have (and the only meter I have that has a capacitance setting) is an Extech EX420. The Extech meter can't seem to detect any value at all with either side of the HVAC cap, it shows a change (or a difference using the range function) of .08 µF and .09 µF for the two values. Which looks like "a margin of error equal to nothing". Since the motor acts like the cap is bad, this would make sense on its own.
At this point I realized that I couldn't remember ever using the capacitance setting for anything, so I decided to test it against some new presumed-good capacitors. I grabbed three different electrolytics of 100, 47 and 4.7 µF and tried checking them. The EX420 would not register values at all when used the normal way. If I used the [range] button to zero-out the stray capacitance of the leads, it would sometimes hit almost right on the labeled values, but other times it just says "OL". It seems to take 3-4 seconds to show anything on the display, which is really rather slow compared to how fast it can show volts, amps or ohms. (-I leave it on the auto-ranging setting usually-)
So then I tried a couple different values of ceramic caps--and the same meter (without using the range setting) hit right on the labeled values, in maybe 1/4-second. It might waver 1 or 2 counts, but it quickly settles right on, or very close, to the stated values on the caps themselves. So it measures ceramic caps correctly, and just as fast as it can show volts, amps or ohms.
Investigating the matter further online let me to the subject of the problems of testing caps, and of ESR testing for electrolytic caps, that I'd never heard of before.
I have found the cheaper meters that do this, the "auto-detect" transistor testers based on an ATMega328. They usually only have a 14-pin ZIF socket and a [test] and [off] buttons. There is open-board ones for ~$10-$15 around, and a couple in a rather generic looking cases for ~$25 or so.
What I am wondering here is, is there any device better quality than that? There's a few "transistor testers" with only 3 leads that claim to do ESR testing, but they only have a digit readout and no buttons to select a mode, so I don't understand how that would work.... I've looked at a few lower-cost multimeter-style RCL meters, but none of them seem to have the ability to measure ESL. I'm thinking that there must be a better/higher-quality machine for doing this, but I can't seem to guess what it's called.
I have also seen a few pages where the opinion is that since you can't test a cap properly without removing it from a circuit anyway, that it's usually just easier and faster to replace it than it is to cut it out, test it and put it back in if it's good. The principle of the situation annoys me though--that you can have this part and you have NO way to really tell if it's working or not.
Do you have anything to test ESR with? Is it one of the above two (cheaper) ATMega328-type things, or is it something better?
....The funny part about this situation is that just TWO DAYS AGO, I decided to order a couple new DMMs, just because the one I tend to use the most out of habit is pretty old and rather inexpensive nowadays. And both of the new ones also have a capacitance function, tho neither measures ESL... but I'd at least have had another different meter to try testing with. :|
The capacitor in question is a round-can 440v 40/5 µF castor-oil-filled capacitor from an outside condenser of a central-AC unit. The indoor AC unit blows air, but it isn't cold.... Casual observation showed that the condenser unit fan will not start on its own, but it spins freely and will run if I spin the fan with a stick--and then the AC produces cold air like normal. The outdoor condenser unit was replaced only ~4 years ago, so I am guessing the cap just failed.
After shutting off the power to the AC system, I removed the capacitor and will just go shopping for a new one tomorrow. I got curious tho and decided to try testing the capacitor, since I had one DMM that could do that (I thought).
The best meter I have (and the only meter I have that has a capacitance setting) is an Extech EX420. The Extech meter can't seem to detect any value at all with either side of the HVAC cap, it shows a change (or a difference using the range function) of .08 µF and .09 µF for the two values. Which looks like "a margin of error equal to nothing". Since the motor acts like the cap is bad, this would make sense on its own.
At this point I realized that I couldn't remember ever using the capacitance setting for anything, so I decided to test it against some new presumed-good capacitors. I grabbed three different electrolytics of 100, 47 and 4.7 µF and tried checking them. The EX420 would not register values at all when used the normal way. If I used the [range] button to zero-out the stray capacitance of the leads, it would sometimes hit almost right on the labeled values, but other times it just says "OL". It seems to take 3-4 seconds to show anything on the display, which is really rather slow compared to how fast it can show volts, amps or ohms. (-I leave it on the auto-ranging setting usually-)
So then I tried a couple different values of ceramic caps--and the same meter (without using the range setting) hit right on the labeled values, in maybe 1/4-second. It might waver 1 or 2 counts, but it quickly settles right on, or very close, to the stated values on the caps themselves. So it measures ceramic caps correctly, and just as fast as it can show volts, amps or ohms.
Investigating the matter further online let me to the subject of the problems of testing caps, and of ESR testing for electrolytic caps, that I'd never heard of before.
I have found the cheaper meters that do this, the "auto-detect" transistor testers based on an ATMega328. They usually only have a 14-pin ZIF socket and a [test] and [off] buttons. There is open-board ones for ~$10-$15 around, and a couple in a rather generic looking cases for ~$25 or so.
What I am wondering here is, is there any device better quality than that? There's a few "transistor testers" with only 3 leads that claim to do ESR testing, but they only have a digit readout and no buttons to select a mode, so I don't understand how that would work.... I've looked at a few lower-cost multimeter-style RCL meters, but none of them seem to have the ability to measure ESL. I'm thinking that there must be a better/higher-quality machine for doing this, but I can't seem to guess what it's called.
I have also seen a few pages where the opinion is that since you can't test a cap properly without removing it from a circuit anyway, that it's usually just easier and faster to replace it than it is to cut it out, test it and put it back in if it's good. The principle of the situation annoys me though--that you can have this part and you have NO way to really tell if it's working or not.
Do you have anything to test ESR with? Is it one of the above two (cheaper) ATMega328-type things, or is it something better?
....The funny part about this situation is that just TWO DAYS AGO, I decided to order a couple new DMMs, just because the one I tend to use the most out of habit is pretty old and rather inexpensive nowadays. And both of the new ones also have a capacitance function, tho neither measures ESL... but I'd at least have had another different meter to try testing with. :|
