EEVblog 1474 - Can You Measure Capacitors IN Circuit? 

for sure you need to understand what you are getting for your results and know why you are getting those readings by studying the circuit !

How can these be measured in circuit especially if there are caps in parallel?

​@@irishguy200007basically it just adds them up but esr should be the same. For example if you had three 100uf caps in parallel your meter will read it at 300uf, just have to look at the circuit the go from there.

In short, it's a trap for young players. When you're a gray beard and know all the limits of your measuring equipment, it'll be fine. When you are young, and don't know what the circuit is doing, and don't know what your multimeter is doing, and don't know what good vs bad looks like, it won't be fine. So just like everything in electronics!

Totally agreed, I'm getting one now for my toolkit! Most DMMs have a capacitance function but you'd have to remove it from the circuit.

@@TradieTrev FWIW, when you are intentionally measuring capacitors on a board to troubleshoot, the capacitance is often the *least* helpful of the values to rely on as most capacitors have to be in extremely poor shape before they tend to vary much in capacitance value. That makes that capacitance function in a typical DMM almost useless for troubleshooting unless one is just trying to learn the rough capacitance of an otherwise unlabeled component. The reason LCR meters are so valuable as a troubleshooting/diagnostic tool is that they measure ESR among their many parameters and that is typically the most critical diagnostic value in the majority of troubleshooting applications when dealing with capacitors (especially electrolytic) as the ESR will rise LONG before the capacitance begins to suffer appreciably. When you see a 1000uF capacitor that still measures as 1000uF and looks absolutely fine visually but has an ESR of 20-ohms you *know* you found a dry electrolytic capacitor that urgently needs replacing and is likely the cause of the failure. Yet an ordinary DMM would still lead one to think that the capacitor was good because it still measures the expected 1000uF capacitance…and that can easily lead one astray when starting out. In most cases, there will often be other marginal or failing electrolytic capacitors that also need replacing due to wear and tear from years of heat and runtime by the time the first actually dries out enough to cause the device to misbehave (barring very poor design or crap capacitors!). Being able to measure the ESR value with an LCR meter can really facilitate locating the failed capacitor(s) in a hurry. About the only exception to that rule is film capacitors (such as X/Y types, Mylar, polypropylene, etc.) as their failure mode is typically momentary shorts blowing holes in the film layer to clear shorts where the dielectric has failed (usually due to AC mains voltage spikes), which is an intentional part of their design for safety in certain applications. Since the faults actually blow holes in the conductive layers of film capacitors, consuming plate area, they will often show appreciable loss of capacitance well before a significant rise in ESR, but that is largely the “exception that proves the rule” in troubleshooting and that is something you learn to lookout for when dealing with film type capacitors versus electrolytics. If you are already thinking about getting an LCR meter and have more than a mere passing interest in electronics and repair of modern and vintage electronics then an LCR meter is a great investment after a good DMM as so many faults in consumer and vintage electronics are the result of failing capacitors and a few dollars in decent quality capacitors is often all that is needed to get things working again. The DE-5000 LCR meter can be found at a very good price (the iET version costs a fair bit and commands a premium, but if you don’t need any certification or the minor enhancements the iET model has then the knockoff DE-5000 by DER EE is practically as good and available at a very favorable price (around $100 USD, give or take, sometimes much less used) for an excellent LCR meter. Just remember to always discharge capacitors before you connect them to your LCR meter as a capacitor with a high enough voltage charge can damage an LCR meter, so it is good practice to short the capacitor (ideally with a few tens or hundreds of ohms to avoid welding your shorting probes to capacitors with especially high voltages and capacitances) immediately before connecting them to your LCR meter. Of course, if the board has not been powered on in ages then you can skip that step as can you if you know for certain that there are no voltages anywhere on the board that exceed the safe maximum voltage that your LCR meter can safely handle (which should be listed in its documentation), but discharging capacitors first is a good practice to avoid accidentally damaging a perfectly good LCR meter because we regularly find ourselves testing mains voltage capacitors in power supply circuits as they are one of the most common faults. Good luck and I hope you enjoy learning to use your new LCR meter when you get one! They are a great piece of kit to add to your test gear and bench!

@@ethanpoole3443 I'm sure OP and myself couldn't agree more with your input. Typically caps are my go to thing when troubleshooting power supplies.

​@@ethanpoole3443 I did hear the DE5000 won't do in circuit measurings? Still looking for a fairly reasonable priced esr meter Which can do in circuit measuring

Yep, forgot to mention that.

Even better, 4W plus a guard connection

Yes, because people are used to seeing OL as OPEN on their multimeter ohms range

@@EEVblog My TPI Scopemeter says OUCH on overload.

Over limit makes sense in both cases...?

@@peterlarkin762 Some people claim it's definitely open line and won't hear anything else.

I have the purple one, made by Peak electronics, atlas esr model esr60 the manufacturers claim that it will measure capacitance in circuit. Certainly seems to do that without much problem.

is it worth it? i have 3 peak meters, semiconductor analyser dca 75, the scr and diode tester models, i had choosen the de 5000 over peak lcr because it can do more

@@Equiluxe1 On the Peak Instruments page for the ESR70 (the current model of purple bean) they say it will only measure the resistance whilst in-circuit. Out-of-circuit it will measure both the resistance & capacitance. For their two models of LCR meter (the yellow beans), they specifically state that they don't recommend using them for in-circuit testing, but this is mostly just them covering against claims due to misuse.

BS

@@geirtjelta ?!

in circuit cap containing charge ? someone needs to go back to school .

@@RESISTAGE Many Capacitors in circuit, specially in Switching power Supply, Keep their charge for a long time and they need to be discharged, before you work with them. Here is an example, an AC source gets converted to DC, at say 220V AC, and stored in a Capacitor then follows a Buck converter to step it down to say 12V. When the circuit is turned off the buck converter stop switching and the charge still remain in the cap. They tell you this both in the in circuit capacitor testers manual or safety classes you need to take to work with high voltage circuits. To see this for yourself get a computer Power supply and connected to power then unplug it. then comeback in one hour and measure the voltages on the capacitors, you will see there are sill substantial voltage remains on the cap, which could be dangerous to people who repair these circuit and to to the sensitive equipment.

I will note that the DE5000 manual warns you to discharge caps before measurement. It does not mention whether that's in-circuit or not, but it's right there in the manual. I agree that in-circuit, the cap is unlikely to have a charge, though I guess it depends on the specific circuit...

What brands for analog multimeters that your father use?

@@hwanglee3330 the one I remember was a Creative Electronics brand ESR meter. He had a Simpson analog multimeter, but he mostly used a Fluke 77 DMM.

@@MatthewAlbert-zoonotique thank you to let me know it. Many people recommend analog ESR to do capacitors troubleshooting in circuits. Some source said that it should be less than 1000 pf ( without desoldering).

Cool, I do the same work as you. For most for cal labs , the stuff the manufacturer stopped servicing. But I have repaired a few problem cases the manufacturer could not repair. The only other thing I do next to industrial boards like you do, is board repair for maritime use (but that is most times via the service contractor that normally works on the ships. Those are the boards you can not power-on. I specialize in component desoldering and then testing (I can test almost everything (excepts programmed or custom stuff ) I test because I need to know what the fault is. If I replace all caps without testing and all other components are good I do not know if I fixed it. If those caps were bad the chance is high I fixed it. If they turn out to be good, then the board is not fixed but then I would not know that. I have several ESR meters, also some from my own design, but I do not use them. I have a big collection LCR bridges. (and several modern very good LCR meters) Also some impedance meters and VNA' s. My pride and joy is a General Radio C bridge setup that was used in GR' s own cal lab as master bridge to measure standards caps. It can measure with aF resolution :-) To often there are several in parallel with other caps, but besides that. The biggest problem is that if there is a good one somewhere in parallel a bad one can measure as a good one. I can not afford to take that risk. On the other hand, desoldering is like 10 seconds job so I do not bother to play around with in situ measuring (I seldom have schematics so I can not check that in the schematics) Indeed, also my experience, DC leakage is important. Most times if I find bad caps it is DC leakage, sometimes a tan delta (Dissipation factor) that is outside specs. Caps outside the capacitance specs are very rare, unless D goes through the roof or worse, like in consumer crap, they are popped, but in that case I do not care about any spec :-) Normally the capacitance is more or less fixed by the construction. The aluminium strips do not change in size and if they do not leak the oxidelayer is thick enough. Normally if capacitance is outside specs the DC leakage will also be outside specs. Exceptions make the rule. There are rare things that can happen like a partly eroded/eaten away contacts between foil and pins. Or damage caused by shorts due to overvoltage but in such away it does not permanent short the cap. I often take dead caps apart to better understand the ways they can fail.

@@pa4tim absolutely wonderful and info packed reply! That’s a short summary in master class on most of what I’ve learned over the last 15 years doing component level repair. I very much feel the same about it, and I take the “shotgun“ cap replacement approach most of the time as well. Especially in boards that are for industry or metrology related stuff, not only is it easier but it’s simply a good idea. If something is having issues, and it’s already on my bench, then why in the world wouldn’t I take the time to replace all the electrolytic’s. The only other thing to do would be to go through and meticulously test each one out of circuit, which by that point I would be a fool not to just replace them. And I feel the exact same way about my LCR bridge and component collection. They are like my children. It all started with the Syncor lc153 years ago, and then I stumbled into the world of metrology grade NIST traceable stuff, And the “volt nut” Side of the Internet. It’s incredible the levels of precision one can reach in a private lab these days, short of Josephson junctions and the like… It’s still amazing what is possible. Especially if you work with industrial clients and laboratories… Because they’re always getting rid of equipment that is mind blowing and expensive and useful in a lab like ours. The longer you spend around clients like that, and communicate to them the things that you keep an eye out for and want to be notified of… The stuff starts falling in your lap and you don’t have to look for it as much. I really don’t have an excuse for owning high end calibrators or for my collection of 8.5 digit meters, maybe every now and then I need that level of precision and reliability… But it’s just something I greatly enjoy and admire. To me it’s the ultimate work of art to find a timeless and perfect designs…..Makes me feel like a better engineer just to have that stuff near me while I’m working 😂🤷‍♂️. Hopefully I will absorb some of the greatness😁👍

For sure.... but considering the prevalence of SMD, lifting a leg is not really practical....

@@dannixon247 that's right, gotta desolder it completely for that... or use teeny weeny meter probes to test it in circuit :)

I would like to know also

Thank you for the information

I just have measured on resistor in circuit with a cheap multimeter, and it gave me negative resistance. What was that ??

@@dmitripogosian5084 I've had stuff like that before. Is the circuit powered on? (it should not be). There could also be a residual charge leftover in an in circuit capacitor.

@@bill6255 No it is not power, but yes, capacitor should be involved (the resistor is connected to the ground via 30 muF capacitor). It looks like I am charging the capacitor by measuring, reading are moving while I am measuring, and depend on direction of leads application. The problem is that this part of board is blown (by myself :) and some transistor and 555 timer are soldered out. I am searching of what else may be blown, possible this 1.2 Mohm resistor itself.

@@dmitripogosian5084 I totally thought of the capacitor charging after I posted the reply, it usually charges to the right value. Look for heat damage (like hole blown in side) on resistor bodies or resistors that have be overheated for long period of use. good on you for catching and fixing your self created blown trace. Good luck and please post when you succeed!

Only if you have kelvin connection leads.

@@EEVblog Even the simple clip assembly from the cheap DE-5000 is 4w. This expensive agilent 1733, didn't it come with a similar adapter? That way, ESR measurement, which is always in the mOhm range, makes no sense with such a setup.

Are you sure the 4 wire clip set is properly wired kelvin connection? I have come across a video saying it is not so with the DE-5000 supplied kelvin clips?

@@jayytee8062 my de5000 definitely came with proper kelvin probes - both the short alligator clips and the smd twizzers. I have since desoldered the alligator clips from the adapter and soldered longer kelvin clips in their stead.

@@ASJC27 Okay sounds good. It's just that I've heard while the the 4 wires are there, they are not connected properly for it to be a true kelvin connection. There was a video a while back with some one opening the black box kelvin connector showing this as such.

How much is it?

Unbiased electrolytic caps work Ok with low amplitude AC signals.

A lot of Benchtop LCR meters have bias voltage setting for testing the parameter variations at different voltages.

No problem, here you go. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Mp0n-Gu0aKk.html

Not power but capacitance

There's a fine line between fire and fine. Namely the line extending from the end of the r down to the bottom of the row of text.

Go to Electro Boom's Channel for that!

I'd say increasing the cortex surface area while preserving the size :)

@@KeritechElectronics Son of a Intracranial hematoma! Good answer. I tip my loose fitting hat to you :O) Cheers KS

Wasn't the source AC? Wouldn't matter which way the leads were.

@@xyzconceptsYT *facepalm* yes, of course! I didn't think in AC :D But still, that slightly charges them in reverse polarity half the time. I guess they can take that as long as the voltage is quite low?

@@yesyes_uk Yeah its pretty safe at the low test voltage 👍

@@yesyes_uk I think it’s more that it never spends long being reverse polarized, as the frequency is at least 100Hz. If you put 1V reverse DC on an electrolytic, I’m pretty sure it’s enough to blow it up, but at 100 Hz AC it’s fine.

Oh, and hes incredibly insensitive and ignorant. He constantly complains about the crap he fixes because someone who cared about the device tried to fix it themselves and failed. Not everyone has the same brain as you do, carl, but he cant handle that, so he lives in the basement and spends unreasonable amounts of time trying to fool the planet into seeing him as better than them in some way... hes just rude and his content is so unrealistic its useless.

Don't knock Carl, he might melt.

Hater.

Go to the settings in upper right corner and find “playback speed”, change it to .75. 👍🏼

Randon example: AVX TPS Low ESR Series tantalum datasheet the spec column is literally "ESR Max. @ 100kHz(mΩ)"

@@EEVblog for normal wet electrolytic caps. But the aluminum electrolytic smd cap you measured and then show the Rubicon datasheet state impedance. That datasheet shows impedance at 100kHz while you say it shows ESR. I looked up countless datasheets. I repair calibration and measurement gear for a living, I always look up the datasheets. I have seen ESR for a wet electrolytic cap at 100 kHz once, but that was a huge special HV cap for a 4 kV capacitor bank used to fire a big laser. And indeed for dry tantalum smd caps (like the AVX TPS) the norm is ESR at 100 kHz. But you did not measure those in your video. That is a totally different thing. Those things do not fail often (If used correct) at least, the once I see, maybe it is different in cheap consumer stuff. In the 60/70' s they often used tantalums in early switching power supplies and they fail a lot (but to be honest, often 40 to 50 years later) I have done some tests on solid tantalums and for most the leakage test is interesting. Just a few Volt over-voltage makes them behave very weird. They most times first intermitted fail, sometimes seem to recover partly but in the end they fail. Without current limiting like in my tests they often blow up. Without over voltage tantalums can last forever. To be complete, those aluminum SMD caps you measure soldered and desoldered have a bad name because they used to have liquid electrolyte before the 1990' s. Those where the caps that destroyed many Tek scope. The dry once are pretty durable as far as my experience. I liked your very good video, I watched after this one, about how an LCR meter works so I was a bit surprised you mixed impedance and ESR

@@Okurka. Not my native language, sorry, I think it is fairytale (I hope)

@@pa4tim but at 100 kHz isn’t impedance and esr practically the same since the reactance is much lower than esr at such a high frequency, even with fairly small electrolytic caps? For example, the smallest cap in the data sheet in the video is a 4.7 uF, so reactance is 0.339 Ohm. Impedance is specified as 1.45 Ohm, so the esr is by inference sqrt(1.45^2 - 0.339^2) = 1.41 Ohm.

@@ASJC27 Yes, that is correct.

I'm here for entertainment and education. Most of us are

Drink some coffee and you'll keep up 😉

Holy sh*t, the audacity of this guy 😅

@@offspringfan89 sh#t isn't holy. Only God is holy.