Just wow. snake oil electrolytic capacitor checking method

[paul cotter]

Yes Xena, "can not isolated" rings a bell, I have seen that so many times and I forgot about it.
Axial, to me means lead(s) arising from the end(s). Radial ,means arising from the side, as in a disc ceramic.

[Bud]

Came across electrolithic caps construction docs from TDK and Rubicon. Neither one says anything about caps casing being tied to the negative lead.
Interestingly though, Rubicon says the plastic sleeve is for markings, not insulation. They say if you want casing insulated, speak with them, they have special insulating sleeves options.

[Xena E]

Yes Xena, "can not isolated" rings a bell, I have seen that so many times and I forgot about it.
Axial, to me means lead(s) arising from the end(s). Radial ,means arising from the side, as in a disc ceramic.

OK. But the convention is that a radial electrolytic is one that has its leadouts around the perimeter of the bottom seal not either end at the centre...

[Xena E]

In a few examples of known good small electrolytics, the can behaves as if it is another (leaky) capacitor plate, testing between that and the other terminals.

A faulty one will have to be found and tested.

[Xena E]

OK.
So, I've just asked around about this amongst the knowledgeable at work and one greybeard said he'd heard of something similar.

One of the failure modes that we know about of course is that electrolytic cap will dry out, particularly those of a certain vintage. Testing if there is no connection of any sort to the can in the way of leakage through the electrolyte would mean a faulty component perhaps? 

I think the dude in the first video has just heard some 'old wives tale' and has put his own spin on it.

[paul cotter]

That's much as I expected. The can will have contact with the electrolyte and random voltage may appear on the can but no meaningful interpretation can be made.

[nonius_]

[XenaE] One of the failure modes that we know about of course is that electrolytic cap will dry out, particularly those of a certain vintage. Testing if there is no connection of any sort to the can in the way of leakage through the electrolyte would mean a faulty component perhaps?

That's the mechanism as explained in the Huntron video I linked to earlier in this thread: dried out cap means no (or little) coupling from case to the plate(s). That can be detected using an aanlog signature analyzer. No idea how that could be detected using a DMM (voltage measurement)

I think the guy in the video in the startpost is replacing practically all capacitors - most he 'tests' have a 'negative voltage' on them and thus are 'bad'. After replacing just about all caps in a device it's no surprise that it works better, 'proving' to him his test-method is valid.

He might just as well dispense with his test and replace all caps sight-unseen, and he'd get just as good results, I expect. And save a little time.

But the guy is entitled to his own misconceptions and magical thinking.

[nonius_]

As I had never used the testing method Huntron showed myself, and as I happened to have about 30 bad capacitors lying on the bench after a recently restored Panasonic NV-370 videorecorder, I figured why not give it a shot.

Unlike Huntron I didn't test in-circuit (after all, they had already been removed from the device after being judged defect), and from negative to case. That the tests weren't done in-circuit doesn't matter as we always isolate the caps by testing them on their floating case.

First time I used/experimented with this method (inspired by this thread), but it appears the Huntron-test is indeed valid: it was quite simple to discern good from bad ones.

For example, a bad 33u/10V capacitor with an ESR=17 ohm. The first image shows measured from - to +: a vertical ellipse: (all images 2V/div horizontal; 0.5mA/div vertical)

[ Specified attachment is not available ]
(hm, can't seem to post this image in-line? Anyway, it's still attached at the end of the post; it's not really very relevant for the method at hand, but just illustrates the difference testing pin-to-pin vs. pin-to-case)

But measuring from - to case shows a horizontal-ish ellipse:



Clearly a bad capacitor, as had already been previously established about a week ago using just the ESR-tester

This is what a good 10u/25V capacitor looks like (measuring negative to case):



A bad 10uF/16V cap with ESR=190 ohm:



A bad 33u/25V cap with ESR=44 ohm:



A bad 47u/6.3V cap with ESR=80 ohm:



Compared to a similar but good 47u/6.3V cap:



As can be seen, the difference is remarkable.

The method requires some experience correctly interpreting the results, but the very bad capacitors are easy to spot: horizonal line = very bad. I've so far always relied on my homebuilt ESR-tester, but will definitely use my Huntron-ish homebuilt device in the future too! I'll be testing a few more known good capacitors tonight to be able to better interpret the results. But I'm impressed so far.

[Xena E]

Good work nonius, well done.

Perhaps it would be worthwhile if you could move your post to its own thread so it's not lost in the quack forum of dodgy tech? 😁

I've saved some bad caps to play with as this is interesting, if you can detail your exact method then others can play... I assume you just used a scope for this?

It's good to get to the bottom of the truth behind the legends. 🙂

[nonius_]

The apparatus used was the homebuilt Huntron/Octopus that was shown earlier in this thread: https://www.eevblog.com/forum/dodgy-technology/just-wow-snake-oil-electrolytic-capacitor-checking-method/msg6345730/#msg6345730

Moderators are free to move it to its own topic, but after writing and editing this long reply in the letterbox-size editing area and trying to add the photos in the correct place, wearing out the mouse scrolling up and down, I'm not going to go through that frustration again - once was enough. (trying to resize the editing area didn't work).

[exe]

I can only speculate that there might be some charge from chemical process in the capacitor when dielectric layer got thinner. Not saying this is what's actually happening, it's an example of reasoning.

The method used in the video implies that all caps are connected in a similar way with respect to ground.

BTW, I'm curious if the author of the video tried to measure voltage again, after replacing capacitors.

[Xena E]

The apparatus used was the homebuilt Huntron/Octopus that was shown earlier in this thread: https://www.eevblog.com/forum/dodgy-technology/just-wow-snake-oil-electrolytic-capacitor-checking-method/msg6345660/#msg6345660

Thank you!

Moderators are free to move it to its own topic.

Perhaps they will if you ask. Your findings would be a good subject in the repair forum.

[Mark19960]

Someone sent me another video showing the same technique by the same bloke:



In this case you can see the capacitor is basically free - past 10 seconds you see the capacitor is moving quite a distance which tells me the leads are not soldered.
This cap, like the others, is a DC blocking capacitor that couples the speaker to the TDA2003 audio amp.
That capacitor moves far too much to be soldered in.

[Analog Kid]

It's all bullshit, obviously.