-
Measuring capacitor ESR, what's the best approach?
Posted by
bsodmike
on 24 Sep, 2020 14:26
-
-
#1 Reply
Posted by
TimFox
on 24 Sep, 2020 16:25
-
The DE 5000 is an excellent economical choice for measuring ESR, at discrete frequencies of 100, 120, 1k, 10k, and 100kHz. It is not suitable for in-circuit testing, and can be damaged if connected to a not-discharged capacitor. More expensive lab instruments feature a wider frequency range and better fixtures, but the accuracy of the DE 5000 within its limits is excellent for the price. Be sure to do the open/short calibration with the fixture to be used before measurement.
I don’t believe that IET still sells the 5000, or their tighter-spec 6000 any more. DER EE is the manufacturer, and the DE 5000 is readily available on eBay.
-
#2 Reply
Posted by
bdunham7
on 24 Sep, 2020 16:47
-
The DER-5000 seems a popular choice, I dont' have one so I can't comment further. Tonghui also makes some nice versions and the may be readily available there, I don't know. They're sold under the BK Precision name here, along with the Motec version, which I have. I also have a Mega328 based tester that actually isn't bad for what it is.
You asked about approach, not just the instrument. Just be aware that there won't be a single number for ESR, or even capacitance for that matter. I was refurbishing an old Harrison-style HP power supply and it had this large, short, fat, blue Sprague electrolytic, something like 15mF @ 16V or something like that. It tested normally at 120Hz, even had decent ESR. Just to try it out, I tested it at 100kHz and got a negative reading, which on my meter means that at 100kHz it is an inductor. Every capacitor does this at some point, but you don't usually see it with a meter like that. If you were designing a PS filter, you would test like that and make a curve to see if you need another capacitor in parallel and what it should be.
-
#3 Reply
Posted by
andy3055
on 24 Sep, 2020 17:36
-
-
#4 Reply
Posted by
Peabody
on 24 Sep, 2020 17:49
-
I'd like to know if my method is any good. I connect the positive lead of the capacitor directly to the output of my scope's 1KHz 3.3V square wave test output, and connect the scope probe to the same point. So with little resistance, the capacitor is almost a dead short, but with high ESR, you see a charging curve. I find that I need to set my scope using a known good cap of the same value, then see if the cap under test looks the same. Of course this is only at 1KHz, but seems to work for the electrolytics I deal with.
-
#5 Reply
Posted by
pqass
on 24 Sep, 2020 19:31
-
FYI: How to measure ESR with an oscilloscope and function generator:
-
#6 Reply
Posted by
David Hess
on 24 Sep, 2020 23:00
-
An impedance bridge works well also; I have an ESI 250DA. But a modern LCR meter like a DER-5000 is much faster and easier to use.
There is more to capacitors besides ESR; I have had aluminum electrolytic capacitors which tested fine for leakage, capacitance, and ESR, but were still worn out. A network analyzer will reveal a lot more about what is happening.
-
#7 Reply
Posted by
tggzzz
on 24 Sep, 2020 23:10
-
I've seen how one can measure ESR by using a DMM & signal generator; as I recall the test frequency was around 100kHz @ 1Vpk-pk.
What's a more practical device to use for this, IET DER-5000?
Don't become fixated on one general purpose measurement/parameter, which is probably a theoretical simplification of a complex practical reality.
Do work out what is important in your circuit, and measure that.
-
-
If you want the DER-EE5000, then buy a kit which also has the probes in it. They're very useful, and hard to come by separately.
If I'd buy an LCR meter it would be the DER-EE. It's not the latest nor greatest, but it's a good work horse for a relatively small price.
That said.
Also get one (or two) of those avr uC based "transistor testers" from china.
There are many variants with different displays, but they probably all work and are very handy to have by for all kinds of simple measurements on components. It also tells you the pinout of transistors, and whether they work. It measures capacitors inductors, ESR and lots of other properties.
It's not an accurate meter, but very useful none the less.
-
#9 Reply
Posted by
Psi
on 25 Sep, 2020 10:26
-
I recently bought the ESR70 peak atlas unit to go with my older LCR and DCA units from them.
Works well.
-
#10 Reply
Posted by
TimFox
on 25 Sep, 2020 13:43
-
If you are looking for that unit, the manufacturer is DER EE, but the part number is DE 5000.
-
-
I am using my very old analog HP 4328A milli ohm meter or the Sencor LC103 for reliable ESR measurements.
-
#12 Reply
Posted by
Dacke
on 25 Sep, 2020 19:24
-
I had the DE 5000, a great little LCR meter and was a good price, but unfortunately after a little over 4 years of working fine, it died. I didn't hit it with a charged cap or anything, I just went to turn it on one day and it was completely dead. Did some basic troubleshooting into it but beyond that, I couldn't invest anymore time into it and I needed an LCR meter. Got lucky and I got a good deal on a BK Precision 880 and I've been using that ever since. Around the same time I stumbled into a cheap Sencore LC75 on Ebay someone had up for $85 buy it now, advertised as not working. Turns out most of the contact points, switches, and pots on it just needed cleaning. The Sencore is great for a service and repair tech for quick readings but it is certainly not a precision calibrated instrument (and you can calibrate it yourself) , I mainly kept it for it's ability to measure larger value caps and for the leakage testing. I would still recommend the DE 5000 though, it was a great meter while I had it.
-
#13 Reply
Posted by
ledtester
on 25 Sep, 2020 19:55
-
-
-
-
-
but that is shameless self promotion.
you deserve it !
-
-
There is more to capacitors besides ESR; I have had aluminum electrolytic capacitors which tested fine for leakage, capacitance, and ESR, but were still worn out. A network analyzer will reveal a lot more about what is happening.
Do you know what in particular was worn out with these capacitors? Is it something that can be determined with a performance test, and if so, what would that test be?
-
-
In some applications a very simple method of assessing whether a capacitor's ESR is correct or not is to measure the ripple across its terminals .... If the ripple is less than or equal to the value specified by the manufacturer of the device, the ESR is correct.
The importance of ESR varies depending on the applications in which the capacitor is used.
In some analog applications, (for example, an audio amplifier) the ESR is not important .... To convince yourself of this, just imagine that you have placed a low resistance in series with the capacitor and analyze the influence that this has on the operation of the circuit.
On the other hand, in decoupling or filtering functions, the ESR can be very important, especially in the filtering of SMPS power supplies.
-
-
The DE-5000 is really good but the ESR70/Plus is also very good. Study these two and pick the one best for you - there is a chance you will eventually wind up with both.
-
#19 Reply
Posted by
Shock
on 10 Nov, 2020 10:32
-
The DE5000 LCR meter is affordable and performs a host of different tests including ESR. Has easily modifiable adapters to increase low measurement accuracy and versatility. Can be used in circuit though not specifically designed for it.
For purpose built ESR testing a Bob Parker design such as the Altronics, Anatek or EVB ESR meters. They optionally come in kit form, have schematics and are easily repairable. Specifically if you are doing in circuit testing (i.e. repair work) these meters either come or can be modified with input protection diodes which helps if you accidentally forget to discharge a capacitor.
I prefer the one shown in Daves video below (now sold by Altronics) it's very easy to read, delta the leads and use. The chart is handy on the front as a rough guide and the case is virtually indestructible. With a few modifications it can take standard multimeter test leads, lithium rechargeable batteries and input protection diodes.