ESR meter - In circuit capacitor testing?

[collisr]

Hi there.

I am new to vintage video restoration and looking for a quick method to identify bad capacitors - the ones that aren't leaking/bulging but are causing the circuit to misbehave.

My question is about ESR meters that claim you can test capacitors in-circuit - i.e. not desoldering any legs!. In your experience, is this possible? I am looking for strong indications that a capacitor is bad based on its ESR value - not a 100% accurate ESR reading.

Would a pre-assembled model such as https://www.ebay.co.uk/itm/112510620302 be of value for this type of investigation.

All recommendations and advice welcome

Rob

[sortedradio]

Hi Rob,
Firstly as I'm sure you know an ESR meter will only identify bad electrolytic capacitors. I use an Atlas ESR70 meter, and it is pretty good at in circuit testing. Where an ESR meter won't help is where you have an inductor, i.e. transformer winding, coil etc, or low resistance in parallel with the capacitor under test. In that case lift one leg of the capacitor if possible.

Martin

[GMB]

I would go further by saying that as it measures ESR then it only detects bad electrolytic capacitors that have failed by going high ESR - which in my experience is only a subset of bad ones in general.

They are handy things to have, but not a definitive diagnostic for capacitor problems by a long way.

[mhennessy]

In modern electronics - especially switched-mode power supplies - then capacitor ESR is critical. In many cases, ESR starts to rise and cause problems long before other symptoms (reduced capacitance or bulging). The ESR meter is invaluable on the service bench today.

However, that modern mindset is not applicable to vintage radios in quite the same way. In fact, ESR is rarely an issue. The original capacitors were not low-ESR types - low-ESR capacitors were not really around back in those days - they arrived with the switched-mode power supply. Modern capacitors have to put up with extremely high ripple currents (easily 10s of amps in really quite modest PSUs), but how much ripple current will a capacitor in an old radio see? Of course, ripple current causes heating (I-squares-R), and that accelerates the drying out of the electrolyte, causing ESR to rise (which in turn causes more heating, so more drying. Repeat).

What really matters with old capacitors is leakage. And no affordable capacitor tester measures that to the best of my knowledge - you have to build or rig something yourself.

Of all the capacitors I remove from radios, I normally put them on a tester, just out of interest. They nearly always read OK in the ESR department. In terms of capacitance, they normally read high. That's because of the leakage, which diverts charging current from the capacitor so the capacitor takes longer to charge, which the meter interprets as a higher value. Sometimes, the meter reports "in circuit", meaning the leakage is so high that the meter thinks there's real stuff in parallel with the capacitor. And of course, a shorted capacitor will always have an excellent ESR reading. Occasionally old capacitors do go short, so watch out for that one.

In short, the quick method is experience. Certain brands are "change on sight", others are "guilty until proven innocent". I much prefer to observe the circuit conditions and follow a logical path to a faulty capacitor.

In general, resist the temptation to do "blanket changes". As mentioned, there are "change on sight" capacitors, but a lot of people change perfectly good capacitors that have decades ahead of them, and that's not good. Again, that's different to modern electronics - give me a SMPS that's got more than 5 years on it and I'll blanket-change the caps, and 9 times out of 10, that is the right thing to do - they lead a difficult live in hot and cramped conditions. But that's not true for radios.

The above is based on my experience of transistor sets from the '60s onwards. If you're into valves, then much of the above is still true. Leakage, not ESR.

Hope that helps,

Mark

[Station X]

The OP implies he's intending to use an ESR Meter for vintage video recorder restoration rather than to restore vintage radios.

[Boater Sam]

Sorry, missed that point Graham, how did you arrive at that please?

[Station X]

By reading the original post.

Originally Posted by collisr

Hi there.

I am new to vintage video restoration and looking for a quick method to identify bad capacitors - the ones that aren't leaking/bulging but are causing the circuit to misbehave.
Rob

[MrBungle]

If you fancy spending an hour or to building something, one of these works pretty well (I built one): http://www.eevblog.com/forum/project...-meter-design/

Works in circuit.

As Mark said though; leakage is important. This is very difficult to measure which is why Heathkit capacitor testers seem to go for so much money as they are about the only thing that does this.

[collisr]

Thanks all. That makes sense. I am typically trying to resolve (mostly) power supply problems with VCRs from the late 70s->mid 80s. I guess this is considered "modern"?

I noticed the Atlas ESR70 does not explicitly say it can work "in circuit" with (Elco) capacitors but as sortedradio says - as long as there is not a low-resistance parallel path the meter can easily take, then it should be okay.

Does these assumptions seem reasonable.

Right now I have 3 machines (a video 2000 and 2 Sony Betamaxes) that have various power-related problems and I need to trace along the entire power-path to track down a presumably faulty elco. I have a basic understanding of electronics, but really learning by trial and error. Taking each capacitor out of circuit to test with my DMM would take an age.

R.

[Station X]

I've found the best approach is to check the ESR in circuit and if it looks bad check for paralleled components which may be affecting the result.

[collisr]

Sounds good Graham. I will shoot for that approach. Thanks again

[Station X]

Just bear in mind that a paralleled component may result in what looks like a very low ESR. My ESR Meter measures capacitance as well and looking at both readings generally enables a diagnosis to be made.

[ms660]

Always try and get a copy of the schematic if possible....it can help, you can see any parallel paths from that so you know what to expect when measuring, nice to hear another Betamax or two has survived the crunchotron, used to work on the first model back in the day before they were released in the UK.

Lawrence.

[mhennessy]

Originally Posted by collisr

Thanks all. That makes sense. I am typically trying to resolve (mostly) power supply problems with VCRs from the late 70s->mid 80s. I guess this is considered "modern"?

For a PSU that is switched-mode, then my earlier comments apply - that is a case when blanket cap-changing is a sensible strategy. Don't overlook the baby ones - those are well-known for causing problems.

However, video machines of this era mostly use linear power supplies, not switched-mode. An obvious exception being the Grundig V2000 machines (or, at least the one up in my attic ).

With linear power supplies, the larger smoothing capacitors can fail, and that will be identifiable by hum bars on the picture, or excess ripple on the supply rails as observed by a 'scope. An ESR meter might help to confirm diagnosis, but it's not critical to that diagnosis.

Up the the attic I've a box full of PCBs from various Philips V2000 machines that were scrapped. I use the cards as a source of components for projects. In all the years I've been recycling components from these, I've yet to come across a single faulty electrolytic capacitor. Component quality in this era was very good indeed, and although video machines run warmer than radios, they do survive extremely well.

I've been thinking hard back to the days when I used to repair video machines, and I can only think of one machine that had faulty capacitors (ignoring SMPS faults). Perhaps I've been lucky. This one had a very dark picture with numerous artefacts, and I was completely baffled by it. With no service manual, all I could do was follow the video signal back from the Scart socket with an oscilloscope and see if it improved at any point. That was a good plan, because I quickly found a duff coupling capacitor - the video waveform looked pretty good on the 'scope on one side, but not the other. Replacing that innocent-looking capacitor improved things a lot, but it still wasn't right. Continuing this method, I soon found another, and then another. At this point, I realised that all these failing caps were the same brand and type. Looking across the main PCB, there were many different types, but I decided to just blanket-change all the ones that were identical to the 3 that I'd found so far (only another 5 or 6). This fixed all the problems - all the other types were just fine. I didn't have an ESR meter at the time, but it would only have told me what I already knew. And if I'd used an ESR meter, I think it would have taken me much longer to have found the fault. To be honest, having wanted an ESR meter for years, I have to say that it's one of my least-used bits of test gear...

In short, other components fail too

Time spent reading up on electronics is never time wasted, and fault-finding and restoration is infinitely more satisfying when you can apply that knowledge rather than part-swapping because a computer says so (we're not car technicians - think how long some of them take because they follow a script rather than their instincts ). If you haven't already got one, I'd put an oscilloscope ahead of an ESR meter any day of the week.

[HamishBoxer]

It was only for some reason Sky Box powers supplies that gave false readings in circuit, any other were fine on my ESR meter.

Don't forget to discharge primary side caps before testing.

Must admit re Betamax (Sony) that a blanket change on C7 etc is a good idea as there are some expensive bits of silicon in there.

[boxdoctor]

I built an E.S.R. meter roughly in the style of a circuit in Television magazine in the early '90s. I cannot remember exactly which year and month it was published. It is a very simple and robust design, and mine is still in use now.
As most of these machines test the impedance of the capacitor at about 100Khz, most parallell inductors have a high reactance at that kind of frequency compared to the E.S.R. values expected in S.M.P.S. and similar applications. So in general the meter can give a reasonable indication with the capacitor under test in circuit.
With switch-mode P.S.Us, class D amplifiers, and similar applications I would say they are a more or less indispensible piece of kit for diagnostic purposes.
Dont forget to observe HamishBoxer's caution about discharging capacitors before testing! Tony

[Diabolical Artificer]

I built this meter - https://www.vintage-radio.net/forum/...ad.php?t=54367 with a large dollop of help from David. It's a simple test in circuit tester that I use rather like dipping an elbow in a hot bath before you put your nipper in. So for any bit of kit, it gives an indication of the probable health of electrolytics, rather than a pass/go, good bad. A couple of example's.

Had a look at a Panasonic 70's amp, as usual the main big caps are ok as are most of the others, but several small coupling caps showed as iffy. Took some voltage readings and several readings on various tranny bases were off. Replaced all the 2.2u/3.3u caps with low ESR types

HH 100w guitar amp, several functions not working as expected, ESR meter showed several bad caps, again small ones, in value and size, again coupling caps, replaced these. However the meter showed the main smoothers as ok, but there was terrible hum and the rails read a lower voltage than expected with the reading jumping around. Put them on a homemade tester, they were leaky.

I've found it a useful bit of kit, but the more experience I get, I've found I use it less, but it still is a handy tool.

Andy.

[collisr]

Thanks again everyone.

I can appreciate most of the pros/cons of the ESR meter here, but the validity of using one on switching/linear power supplies I did not understand.

Let me give a specific example of what I'm trying to diagnose:
- I have a Grundig Video 2000 2x4 Super (GB). I have already replaced the 2 X2 mains capacitors (the original ones turned to smoke when I first powered on ;-), but now I have the following observations:
- When I press the PLAY, RWD, FWD buttons, the relevant motors move for about 1/2 second then stop.

My naive assumption is there is a faulty elco somewhere in the secondary side of the power supply. I have the service manual and should be able to trace back each motor power pin to the power board pin. Then I need to dig into the power supply to find the fault.

I thought an ESR meter would help me quickly work backwards from the power board pin testing each elco in the voltage-rail-path in turn until I find the culprit.

Is this a reasonable assumption (the bad elco in the 2nd power board)? and is this the best method of locating the bad component?

I see above that mhennessy said the Grundig V2000 used a switched power supply - I didn't know that - does this apply to my model? And if so, what difference could that make to my diagnosis?


All hints and tips welcome
Rob

[Argus25]

Putting aside any pros & cons on the theory of ESR meters and the utility value of having & using one, if you are going to get one there is really only one type in the world that I would recommend.

This particular meter applies a very small low amplitude pulse to the capacitor under test, good for in circuit testing. It can also be used for low resistor values. I have compared it in my workshop to a number of other types of ESR meter, including ones which use sinewave currents and using other methods with generators & scopes to assess the ESR. I wouldn't use any other ESR meter, this one was beautifully designed and conceived and despite being a kit, its has a very professional painted and screen printed steel front panel and a very solid case and easy to read LED display.

I'm a very critical judge of test gear, I liked these ones so much I bought two of them. It has become an indispensable weapon in the war that I wage against against defective electrolytic capacitors, which are the nemesis to longevity of all electronic equipment. In general I check the capacitor for value, leakage at its operating voltage and ESR to decide if its ok, or not.

http://www.altronics.com.au/p/k2574-esr-meter-kit/


Photo of one attached.

[joebog1]

Is Argus involved with Silicon Chip?

Regards, Joe.