Guys, getting alot of questions on forums and in the real world about motor run capacitors going bad. I did a long write-up on the old GG pages. Probably need to reproduce that here with updates.
Capacitors are electricity storing devices used with A/C motors to help them start easier and run more efficiently. (Some motors have separate start and run capacitors, others use a run cap for both jobs if the needed starting torque isn't too high). Start caps are only used intermittently, so typically don't overheat as a rule unless the relay that controls them sticks. For our discussion I'll be talking about motor run caps.
They have to be precisely matched to the particular motor they were designed for. Hence, all the different ratings. Normal manufacturing processes as they are running thousands of caps a day off the assembly line means that not every capacitor is going to be exactly the rating printed on the outside. So, they design them with a manufacturing tolerance. It used to be + or - 10%, now it's more like 5-6%.
Part of typical HVAC servicing now is checking the run caps. Many techs use an electronic tester that you touch the two terminals on the cap - with power off and leads disconnected - to see what the caps rating is. Compared to the nameplate determines if you replace it or not. So, a 35 MFD cap (according to the old tolerance) allowed for a 3.5 MFD drop and still be acceptable - i.e. a measured rating of 31.5 MFD or higher would "pass". (I've never seen a capacitor fail high on amps, nor have I ever found one above nameplate rating, so, for this discussion I'll ignore that possibility).
One of the main reasons for run caps "losing" capacitance over time is heat. There's something called "back EMF" in an electric motor, part of the design characteristics whenever motors are used. The back EMF or voltage is much higher than the applied (forward) voltage. A 110 volt window AC might develop 250 volts back emf, a 220 volt residential unit could develop 400 volts back emf. The run capacitor(s) is exposed to this voltage while doing its job, and is usually filled with a thermally stable oil to absorb this heat produced and maintain the insulative qualities.
But testing an electrical device statically like this might not be the most accurate way. I mean, it's going to be under load during all of it's operating life - shouldn't we check it under those conditions? I started doing it, and to my surprise I found a substantial number of run caps that were under rated while loaded and running that PASSED the static test.
Just last week I taught a younger tech this method. He tested the dual run cap (nameplate had 35 MFD listed for the comp part) with his tester and got 33.4 MFD. Not bad - he was going to pass it. I said, lets try the loaded method. Dang if the thing only read 28.6 MFD while loaded! So, he R&R the thing.
Heck, the nameplate said it had a 6% tolerance - that's a 2.1 MFD acceptable drop, for a minimum of 32.9 MFD to pass. We were dang close to that (33.4 MFD) unloaded!
Here's a article that outlines the procedure and reasoning:
https://hvacrschool.com/testing-run-capacitors-smart-easy-way/https://www.achrnews.com/articles/135163-testing-the-run-capacitor-while-the-system-is-running