Building a simple ESR meter for troubleshooting electrolytic capacitors.
"Never in the history of mankind, have so many boards been ruined by so few components"
Winston Churchill (or someone like him) - on Electrolytic Capacitors
Electrolytic capacitors are a very common cause of failure in electronics equipment. Over time, electrolytic capacitors tend to dry-out and fail. Often the failure can be found by simple visual inspection (an exploded or deformed capacitor is hard to miss). In other cases however, the failure is more subtle and manifests itself by an increase of the capacitor's Equivalent Series Resistance (ESR). An ideal capacitor of course, has zero equivalent series resistance. In reality, a multitude of non-ideal characteristics can be modeled as a simple series resistor. Manufacturers often specify ESR at a fairly high frequency of 100 KHz and it is typically a very low value, depending of course on the capacitance value and voltage rating. Typically, the higher the capacitance, the lower the ESR. Large capacitors tend to be used for power supply filtering where even small values of ESR translate in increased internal power dissipation and premature capacitor failure. If you repair electronics equipment, an ESR meter is an essential item in your toolbox. Of course you can buy already made units (See this one from Amazon.com for example) but building one yourself is relatively simple and rewarding. In this article, I'll show you how I built one using common components in my hobbyist Electronics Lab.
An ESR meter is in effect an AC Ohm-meter designed to detect very low value resistors. This can be achieved by applying a fairly high frequency AC signal to the capacitor (100 KHz in this design) and measuring the effects of the resistive drop across the capacitor. At 100 KHz, the reactive component of the capacitor (Xc = 1/(2*pi*f*C) ) is negligible for most common electrolytic capacitor values. The circuit shown in Figure 1 uses this operation principle. A simple 555 timer generates a square wave at about 100 KHz. The 2:1 step-down transformer reduces the effective source impedance of the 555 output driver and couples the signal to the capacitor under test and the two 12 Ohm resistors. The 1N4007 diodes protect against capacitors that might still be charged (in this case, the diodes will effectively "discharge" them to about 0.6V). The resulting signal at point (1) is about 200 mV pk-pk. This voltage is small enough that the tester can sometimes be used in-circuit without forward-biasing any PN junctions. Note however that often several capacitors are connected in parallel in real-world designs so the measurement might be misleading. A 2N3904 transistor amplifies the signal coupled through the capacitor with a gain of about 10. This signal is the rectified, filtered and applied to a simple analog voltmeter. Note that because of the way the capacitor is connected, the voltage will be maximum for a 0 ESR capacitor and minimum for a large ESR capacitor. Therefore, the meter is adjusted by first shorting the test leads and ensuring the meter deflects till the end. This is done by adjusting the series 10K Ohm pot. The scale is very non-linear so you might want to test with know resistor values and mark the corresponding numbers in the scale. I didn't do this in my design but instead kept a separate table that I later glued on the ESR box.
Figure 1 - ESR Meter Circuit
UPDATE: Reader and eevblog fellow forum member "Radio Tech" was kind enough to re-draw in software my original hand-drawn schematic. Here's his version:
UPDATE1: Earlier revision had C5 and C6 marked as 100uF. The correct value is 100 nF (as in the original drawing above).
The circuit is powered from a single 9V battery. To maximize battery life as the voltage drops over time, I decided to use a low-cost LDO instead of a more common 78L05 regulator. The LM2936 is relatively inexpensive and will ensure operation even when the battery approaches 5.5V. Though not shown in Figure 1, I later also added an LED ON/OFF indicator after the switch.
To better illustrate how the circuit works, let's have a look at the waveforms at points (1), (2) and (3) marked in Figure 1 with a shorted lead (ESR = 0) and with a relatively high ESR = 5.6 Ohm. The waveforms (1), (2) and (3) in Figure 1 correspond to Channels 1, 2 and 3 Figure 2's oscilloscope shot. Figure 2 shows that the signal at the output of the capacitor/resistor divider is only about 232 mV pk-pk and that the transistor effectively multiplies that value by a factor of over 10. Channel 3 is the already rectified and filtered signal, in this case about 1.35 V.
Figure 2 - Waveforms with shorted test leads (ESR = 0)
Figure 3 shows the meter I used in full-depletion for this situation (after adjusting the 10K potentiometer).
Figure 3 - Full depletion (ESR = 0)
When a high ESR capacitor (in this case 5.6 Ohm) is inserted across the test leads, the output voltage at point (2) decreases and so does the resulting DC level at point (3) as depicted in Figures 4 (oscilloscope) and Figure 5 (meter).
Figure 4 - Waveforms with ESR = 5.6 Ohm
Figure 5 - Depletion with ESR = 5.6 Ohm
The circuit is simple enough that I decided to use a perf board for assembly. Figure 6 shows the result. I already owned an adequate signal transformer in my parts bin though this is not exactly easy to find (It's used for T1/E1 transmission links). A more common alternative is to use transformers found in switching power supplies such as PC ATX supplies. A larger turns ratio can also be used though this will require adjusting the 150 Ohm series resistor to a higher value.
Figure 6 - Circuit Board
Figures 7 and 8 show the final assembly inside a metallic box. Noticed I used an audio jack (and shielded audio cable) to connect to the test probes. The shield (ground) connects to the metallic case.
Figure 7 - Box and Insides
Figure 8 - Final result
Building an ESR meter is a simple and fun project. Using only commonly found parts around an Electronics Lab, you should be ready to find those nasty high-ESR capacitors in no time!
Comments, questions, suggestions? You can reach me at: contact (at sign) paulorenato (dot) com