testing cap top

Finding Bad Electrolytic Capacitors

testing cap top

Testing capacitor top with test probes (click to enlarge)

As many of you may already know, electrolytic capacitors are one of the more common failures on electronic circuit boards. They have a hard life filtering all of the AC noise out of the DC voltage lines on the PCB. Over time they degrade and eventually the equivalent series resistance or ESR becomes too high to effectively do their job. Methods for finding a degraded electrolytic capacitor (referred to as “cap”) range from measuring the AC ripple or noise across the cap on a live PCB to removing the cap from circuit and using a capacitor checker. Measuring the voltage on a live board may be difficult especially if you do not have the testbed to get the board to a powered state. A cap checker or ESR meter is fine but may not provide accurate or repeatable measurements with the cap in-circuit.

The easiest way to test an electrolytic cap in-circuit is to probe from the metal top of the cap (as shown in the image to the right) to the negative connection (typically ground) using an analog signature analysis (ASA) instrument or DMM that can test capacitance. This works because the metal top of the cap has an electrical connection to the electrolyte material inside but does not connect to the either of the device pins. Testing the top gives you an isolated test point so when you see an incorrect reading then you know the cap you are testing is bad.

Testing to the Top with a DMM
When using a DMM for cap testing you will set the DMM to test for capacitance and allow for AUTO range. The image below shows the comparison measurements of a good versus bad capacitor. The measurements were captured in Huntron Workstation software to provide a good visual comparison. The good 1000uF cap measured 2uF (the “Hold Reading”) to the top and the bad 1000uF cap measured 0.14uF (the “Reading”). Using a DMM may require multiple measurements to get a stable reading since the cap may pick up a slight charge from the DMM voltage.

DMM cap test

Measurements to Cap top using a DMM

Testing to the top will not provide the actual capacitve value (1000uF in this case) since you are not measuring across the device leads. The value you measure is used only for comparison of good versus bad.

Testing to the Top using Analog Signature Analysis (ASA)
ASA instruments such as the Huntron Tracker are well known for making cap testing very easy. They provide a visual representation that is easy to interpret and compare. Since they use an AC sine wave as the test stimulus, the charging effect of the cap is minimized providing a fairly stable reading.
Start by setting the Tracker range to a higher resistance such as 10Kohms and the frequency to a lower value around 60Hz. When you test the top of the cap you will see an elliptical response caused by the voltage/current phase shift inherent to capacitive devices. Good electrolytic caps will show a signature that is vertically oriented which means the device ESR is fairly low. Bad electrolytic caps will show a tilt to the signature indicating that the device has become resistive and its ESR is high. The image below shows signatures captured in Huntron Workstation with the good (in green) versus bad (in blue) comparison.

good versus bad cap comparison

Good versus bad cap signatures using ASA

The difference in signatures is very clear and may be easier to interpret than the numeric values provided by a DMM or capacitance checker.

Final thoughts…
Based on experience, we have found that testing to the cap tops to be the most effective way to detect degrading electrolytic capacitors. In many cases, the cap may still function okay but show signs of degradation in which it is probably best to replace the cap while you have the PCB in hand. Replace the suspect caps now and avoid having to do it at a later time.

Thanks for reading!

Share

3 thoughts on “Finding Bad Electrolytic Capacitors

  1. WTB
    Want to buy

    Im looking for a part for my Huntron 10005B-. It is the 5 pin one stud Oscillator.
    Please contact me if you have one.
    Thanks
    Garry

Leave a Reply

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