Testing Run Capacitors the Smart (and Easy) Way

When testing a run capacitor, many techs pull the leads off and use the capacitance setting on their meter to test the capacitor. On a system that is not running, there isn't anything wrong with this test. However, when you are CONSTANTLY checking capacitors as a matter of regular testing and maintenance, that extra step of pulling the connectors off can be time-consuming. In these cases, it is also totally unnecessary. Testing the capacitors UNDER LOAD (while running) is a great way to confirm that the capacitor is doing its job under real load conditions, which is also more accurate than taking the reading with the unit off.

First, if you are used to doing capacitor checks during the “cleaning” stage of a PM, you will need to change your practices and do your tests during the “testing” phase. You will take these readings at the same time you are taking other amperage and voltage readings during the run test.

This method is practical and is a composite of two different test practices combined –

1. Read your Volt (EMF) and Amp (Current) readings like usual and note your readings.

2. Measure the amperage of just the start wire (wiring connecting to the start winding). It will be the wire between your capacitor and the compressor. When it comes to 4-wire motors, it will usually be the brown wire, NOT the brown with a white stripe. Note your amperage on this wire.

3. Measure the voltage between the two capacitor terminals. For the compressor, that would be between HERM and C; for the condenser fan motor, that would be between FAN and C. Note the voltage readings.

4. Now, take the amp reading you took on the start wire (wire from the capacitor) and multiply by 2,652 (some say 2650 but 2652 is slightly more accurate). Then, divide that total by the capacitor volts you measured. The simple formula is Start Winding Amps X 2,652 ÷  Capacitor Voltage = Microfarads.

5. Read the nameplate MFD on the capacitors and compare it to your actual readings. Most capacitors allow for a 6%+/- tolerance. If it's outside of that range, then replacement of the capacitor may be necessary. Always double-check your math before you quote a customer. We need to make sure we are accurate when advising a repair.

6. Repeat this process on all of the run capacitors, and you will have assurance whether they are fully functional under load or not.

7. Keep in mind that the capacitor installed may not be the CORRECT capacitor. The motor or compressor may have been replaced, or someone may have put in the wrong size. When in doubt, refer to the data plate or specs on the specific motor or compressor.

If you need a visual, we've linked some good videos on the topic at the end of this article. Note that some will use 2650, some 2652, and some 2653. It all depends on how many decimals of pi they are using in their calculation, but all of them will result in an accurate enough conclusion for our use.

At first, doing it this way may take a few minutes longer. Still, in the long run, you will go quicker, have fewer mistakes (forgetting to put the terminals back), have a better understanding of how the equipment is operating, and get a more accurate reading.

Once you replace a capacitor, always recheck your readings to ensure the new capacitor reads correctly under load.

It is also a good practice to check capacitors you have removed with your capacitance setting on your meter as a reference point.

While this method is good, it is only as good as your tools and your math. When in doubt, double-check… and always be in doubt.

—Bryan

7 responses to “Testing Run Capacitors the Smart (and Easy) Way”

  1. As usual, an A+ procedure details. Excellent step 3 comparing Start amps to Run + Common amps – makes electrical sense albeit some reactance impedance impact. However, being kin spirits as a stickler for details and accuracy (if not flagged by others an you placed to see who was closely reading ☺), the step 6 equation narrative has an error but you have correct in actual equation. “…..Now take the amp reading you took on the start wire (wire from the capacitor) and multiply by 2,652 (some say 2650 but 2652 is slightly more accurate) then divide that total by the capacitor amps you measured. the simple formula is Start Winding Amps X 2,650 ÷ Capacitor Voltage = Microfarads”. Thanks for your time sharing your knowledge and experience.

  2. It’s 2653, because of E = I * Xc.
    Capacitive Reactance (Xc) is calculated like this:
    Xc = 1 / (2 x pi x hz x farad)
    Xc = 1 / (2 x 3.1415 (pi) x 60 (hz) x farad)
    Xc = 0.0026527 x farad
    farad = 1,000,000 microfarads
    So 0.0026527 must be multiplied by 1,000,000
    Xc = 0.0026527 ✕ 1,000,000
    Xc = 2652.7
    E = I x 2653
    Also, using this method can make good capacitors look bad if your amp clamp doesn’t have a resolution of .01. Also allow 10% on capacitors to account for meter error.
    Always test both ways using a good DMM before charging a client.

  3. Very informative! I had a question about a “trick” my former boss told me. If I were on an old condensing unit with a bad capacitor…but the model number of the unit is long gone…the compressor numbers are gone…and the capacitor numbers (obviously) cannot be read. I know by size that it is probably between a 30 and a 45 capacitor. I was taught to check start winding amps on several different sizes…..and use the one giving the lowest amps. It seemed good advice…..is it correct? Thanks

  4. I had a question about step 7 – what if I were on an old compressor…..data plate way past being able to read. Same with condensing unit model number…..unit old but cooled good until capacitor went out. I had a boss teach me in this case to test start winding amps on several different sized capacitors which would cover the proper size (25uf up to 50uf or whatever)…..and use the capacitor which gives the lowest start winding amp draw. Is this sound advice? Thanks

  5. This formula has always bedeviled me:
    Xc = 1 / (2 x pi x hz x farad)
    Xc = 1 / (2 x 3.1415 (pi) x 60 (hz) x farad)
    Xc = 0.0026527 x farad

    …because I will quietly state that you can’t take something out of the denominator (in this case farads) and suddenly put it into the numerator. 1/(2*pi*60) does indeed lead us to the magic 2652, so I know it works, but 1/(2*pi*60*F) really shouldn’t be the same thing as 1/(2*pi*60) * F. I know it works, but the math here eludes me.

    1/(2 * 3) = 1/6. 1/(2 * 3) does not equal 1/2 * 3

    Not saying it’s wrong. I know the 2652 thing works.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Related Tech Tips

Ball Bearing vs. Sleeve Bearing Motors
One aspect of motor selection that can get overlooked is motor bearings, and it can have big consequences. Sleeve bearings are most common in residential and light commercial applications because they are less expensive and quieter. They don't have rolling “balls” but rather rely on a thin film of oil on metal sleeves. Sleeve bearings […]
Read more
Use Static Pressure Responsibly
The topic of static pressure comes up frequently in HVAC, and it’s sometimes presented as a test that can be all-knowing and insightful about what is happening with regard to airflow in a system. To use static pressure responsibly, we need to understand what it really is and what it means for an HVAC system. […]
Read more
Condensation - It Isn't Where Hot Meets Cold
Photo by Stephen Rardon I hear it all the time. Someone will talk about undesired condensation on an air handler cabinet, on a supply air duct, in a ventilation duct, or on a vent like the one above. In reply, someone will inevitably say, “Condensation occurs where hot meets cold.” Early on in my career, […]
Read more
loading

To continue you need to agree to our terms.

The HVAC School site, podcast and daily tech tips
Made possible by Generous support from