How does the MrCarlsons lab capacitor leakage tester work without high voltage?

[Beamin]

I understand he does this for a living and has a patreon subscription but I can't afford that as my funds are devoted to medical stuff.

So obviously I don't want people to take the design and post it on the internet. But is there another product that does this that I could google? Or is this a special technique that was invented when things became solid state/safer? I have a feeling it uses radio/AC. Come to think of it I might have a Chinese ebay capacitance tester I could scope once I get my stuff out of storage. The scope unlocks most of the mysteries hidden on a pair of leads. I still have the phobia of "Will this blow up my scope?" though.


Once my situtation changes along with a few other things his page this page, and two other are going to be on my list to support. To think I used to give 25% of my disposibe money to charity. But I beleive in karma and it was better then hording it. I still feel guilty watching and not being able to support. The worst thing to do when your finances are in a dip is to get yourself into a bunch of monthly reccuring charges, unfortunately a lot of my fellow country members haven't learned this. Anyways...




https://youtu.be/LhovRIM5xAo

[gslick]

https://www.eevblog.com/forum/testgear/mr-carlsons-lv-capacitor-leakage-tester!-it's-finally-out!/

[maxwell3e10]

This is a silly product and provides no quantitative information. To measure capacitor leakage current or dielectric absorption one can just build a simple electrometer using  LMC6001 or similar op-amp with fA bias current. As someone pointed out on another thread, capacitors can also develop leakage only at high voltage, so this is not a complete test anyway.

[innkeeper]

I think its a cool concept. the problem with it is, most of the usage of it and understanding the results are based less on measured results and more on inherent knowledge of the expected behavior of the device against comparative results of known good components.

if the functionality could be put into the $20 lcr meter for example and add that inherent knowledge into it and display results that are pass/go and/or numeric form that be great... this could be done by comparing the labeled capacitance and expected capacitor type characteristics perhaps against an expected result table and then displaying a result/results. this would give a more user-friendly device with results you could quantify.

I am on the fence about if HV is needed to detect leakage. i don't doubt his success with the device. though HV is needed to detect issues with breakdown voltage for sure.

with any test gear though, there will be issues that can go undetected no test device is perfect for all conditions. I have no issues with someone building a better mousetrap, and sure, along the way, your gonna let some mice get away

I respect his desire to promote his channel and promote his patron. I do wish though he would publish the concepts and open them for further development and refinement.  I certainly like the direction he is going with the concept.

[alpher]

I understand he does this for a living and has a patreon subscription but I can't afford that as my funds are devoted to medical stuff.

I quite frankly find it hard to believe someone living in the USA not being able to afford 2 bucks
But nonetheless if you're short I'll be willing to cover the cost of let's say a 6 month subscription to his patreon thing.
I'm sure that would answer your questions much better.
Shoot me your email and I will pp the funds.

[CopperCone]

mr carelson has a phobia of lawyers, that's the only reason that product is bizarre. he reviews a real tester.

I also think that secretly he does not want people to reform capacitors using his circuitry, which I agree with. It's like keeping clunker cars on the road.

before you argue about clunkers or garbage capacitors, , I don't want to pay medical bills for your getting shocked or breath the shit those cars produce 

[jolshefsky]

As far as I understand, capacitor leakage is just a parallel resistance that develops over time in nearly all capacitors sooner or later. That eventually makes capacitors act more like resistors. The guts of the circuit is just a detector for very low current which is supposed to go to actual zero in a perfect capacitor when a DC voltage is applied. The lights and all that are just a display of that information.

Note that his specialization is in antique vacuum tube circuitry, so he's mostly looking to test the leakage in ancient capacitors. When a filter capacitor in a modern circuit leaks current, it'll mostly kind of work at low voltages, and ultimately just cause the circuit to stop working. When you're dealing with 600V plate voltages in vacuum tubes, you might burn out an expensive, hard-to-replace tube (valve), or maybe even cause some real fiery harm.

[joseph nicholas]

I would imagine Paul Carlsons tester uses the standard recommended frequency to test capacitors esr.  I think its 10 kHz.  As to that predictive feature about what will happen in the future.  I cant say.  My hunch is it is most likely some snake oil to drum up interest in his Patreon thing.  After all who can say what will happen.

[bitseeker]

Note that this low voltage leakage tester is not an ESR tester. ESR is series resistance and leakage is parallel resistance. They are different properties that are useful for determining the health of capacitors.

[Beamin]

This is a silly product and provides no quantitative information. To measure capacitor leakage current or dielectric absorption one can just build a simple electrometer using  LMC6001 or similar op-amp with fA bias current. As someone pointed out on another thread, capacitors can also develop leakage only at high voltage, so this is not a complete test anyway.

Yes I don't like the lights a 7 seg timer and numeric display would be better then the LEDs. If you're in a rush you might get incorrect results or if you read a value wrong etc. There is no way to know when it's done and not user friendly. I think the box and sparse lights make it look shitty but I think he's trying to make it look like equipment from the magic eye era; heavy clunky and those jewel lights are the worst plastic looking things. I just remember as a kid anything with those fake jewel lights was usually outdated, obsolete and usually burnt out. The late 70's home stereos were the best looking electronics. It's like he's trying to make it look shitty and outdated.

 Some of his other projects are absolute art. He built a box that he wet sanded and clear coated that just hides in his air compressor in his garage. That's a fine line between perfection and pathological OCD. You wont find duct tape in his house.

Still no closer to seeing how this works though. 

[Beamin]

mr carelson has a phobia of lawyers, that's the only reason that product is bizarre. he reviews a real tester.

I also think that secretly he does not want people to reform capacitors using his circuitry, which I agree with. It's like keeping clunker cars on the road.

before you argue about clunkers or garbage capacitors, , I don't want to pay medical bills for your getting shocked or breath the shit those cars produce 

That cappuccino latte coming out of the engine is when oil and coolant forms an emulsion under high temp/pressure and turns into that when the pressure is released. Incredibly damaging to the engine. If you see drops of coffee with cream condensation on the oil or coolant cap when you look at the underside you have big problems. GM engines were notorious for the head gaskets going bad and shooting oil into the coolant. Stay away from buying used cars with this as the owner probably didn't notice and fucked up the engine, or did notice, got an estimate, and that's why they are selling the car.

[stevelup]

To properly test a high voltage capacitor, you need to apply a high voltage to it otherwise you simply don't know what other 'breakdown' mechanisms are taking place.

His tester looks great for testing stuff only exposed to a few tens of volts, but I really don't see how it can prove anything at 400V for example.

[joseph nicholas]

Im pretty sure he is trying to emulate the function of the Heathkit unit shown in the video as an example of what his thingy does but at lower voltages.  This is his model but since he seems overly concerned with safety he has made a watered down version of the old school high voltage Eico and Heathkit units for the kiddies. 

I have an Eico 950b which is a Wheatstone bridge and capacitance tester.  He keeps mentioning that they are getting more expensive and rare thus a safer replacement.  I applaud this but I doubt it does the same thing as the Heathkit tester in his latest transformer depoting video.

[Beamin]

Im pretty sure he is trying to emulate the function of the Heathkit unit shown in the video as an example of what his thingy does but at lower voltages.  This is his model but since he seems overly concerned with safety he has made a watered down version of the old school high voltage Eico and Heathkit units for the kiddies. 

I have an Eico 950b which is a Wheatstone bridge and capacitance tester.  He keeps mentioning that they are getting more expensive and rare thus a safer replacement.  I applaud this but I doubt it does the same thing as the Heathkit tester in his latest transformer depoting video.

I don't really see MrCarlson as snake oil salesman or a bullshit artist although that is his only video you cant dislike or comment. Maybe people were pissed he didn't explain it like he does with everything else or said it can't possibly be a replacement.


We are trying to figure out what it does. And how something can emulate high voltage without it. It might not be a true high voltage test but that wouldn't make sense as he made that thing specifically for testing those horrible paper and wax caps. If this can't test high voltages the overall goal is to increase safety I don't he would release something that wasn't just as accurate. After all the across the line cap is the most dangerous fault as far fault due to fire, or shock from a hot chassis. Plus he never does anything that is less then what he started with. If it is this is a first.

Maybe it's a huge conspiracy and had this all planned out trying to trick you with his over emphasis on safety in all his other videos. Comments are disabled because he is afraid some one found out, his wife maybe, and she exposed him because she won't get the profits of the device if they get divorced. Or maybe it's such a powerful tester the government is afraid it could upset the balance of the entire Canadian test equipment black market! 

How do the cheap eBay testers work? By putting a 0-10khz sweep signal through them and looking to see what frequencies are blocked vs pass through? That must be how this thing works. But how would the tester know if it doesn't know what value it's supposed to be? The mystery deepens! I personally have been waiting since his first video for him to do a detailed explanation and that not what I got.   

[joseph nicholas]

Beaman, hes not interested in explaining how it works because he wants to monitise his activity on the internet.  Heres the business plan, first you hook em for free, then you stop providing free content, then you get em to pay.  There is nothing wrong with this its just a little anoying because he comes across as a Dudley Do Right, out there looking out for the common man.  As others have pointed out on this form, he seems a little creepy.

I think its significant that he hasnt released a video explaining how the fucking thing works.  I also find it interesting he is not responding here about this tester he is so cryptic about, except if you pay for his channel, and even then I not sure you will find out.

[Calambres]

I wish we can get things calmed unlike the other thread here    I'm really interested in knowing how this works.

[Kalvin]

As someone already pointed out, the circuit https://www.edn.com/design/analog/4364022/Circuit-lets-you-test-capacitors implemented using an op amp and two transistors can also be implemented using a dpdt switch and three (FETs) transistors connected as differential pair and few passive components to be able to measure the difference of the leakage currents. So, what's the fuzz?

[bitseeker]

That's how I understood it. (Disclaimer: I do not have access to the schematics nor a description of the theory of operation of the Carlson low-voltage leakage tester.)

Apply a voltage, measure the current. If it's too high, then the capacitor is leaky. The top row of LEDs indicate how much leakage there is, the range of which is not labeled, but you could probably add your own labeling.

Is it a complete replacement for the high voltage leakage testers? No. The voltage matters to really know if the capacitor is OK for the environment in which it much operate. However, if it leaks at a much lower voltage, then it's clearly a bad one.

If I remember correctly, he said that his tester is a safer and less expensive alternative, rather than a direct replacement, for the high voltage testers. Like any tool, it has trade offs, but once you learn its behaviors and how they change with different types and values of capacitors, it may be sufficient. If not, then grab yourself one of the good old high voltage leakage testers. I found a Sprague TO-6 locally (buried in a guy's garage) for a good price. Get them while you can (or build one).

[Kalvin]

If you look carefully, you may notice that there are two capacitors in the design: Cref and Cx. The Cref is a known good, low leakage capacitor and the Cx is a capacitor to be measured. First, the capacitors are connected in parallel and the capacitors are charged to a test voltage. Then the capacitors are disconnected from the test voltage by the DPDT switch. The voltages of the capacitors are the same, but the voltage of the Cx will start decreasing due to the its leakage. The differential FET-transistor stage will amplify the voltage difference of the capacitors, and the difference can be indicated using a galvanometer or LED bar display, for example.

[IanB]

As far as I can tell by observation, the device applies a voltage to the capacitor and measures the current. The current peaks while the capacitor charges, and then for a non-leaky capacitor declines to zero. If the capacitor leaks the current never gets to zero and the steady state current is indicated on the bar graph. The pass/fail indicator changes color depending on whether the value of the residual current is greater than some threshold.

There are obviously some extra details in the design, such as how to detect different value capacitors and scale the response appropriately, but nothing appears to be very hard to understand in concept.

[MiroS]

I like Mr Carlson  chanel, our mate invested a lot of effort and  $$$  for our fun. I wish to him a financial success in every sense. I am not a subscriber of  payed service of Mr Carlson, electronics for me is now only  a hobby  and I find the fun of  learning myself, so here you are my tester  a bit simpler by list of components 

Tested with 9V and 30V and  1uF, you will find that  higher voltage  better visualizes   leakage current.

[innkeeper]

whats the difference between the "Carson' tester and a resistance test or tests done when reforming a cap...
as part of the reforming process...you're basically charging the capacitor to some known voltage and seeing what the leakage current is at that voltage, or in this case if it ever completes a charge.
then removing the charge and testing self-discharge time which is another indication of leakage.

of course in a testing scenario, you would charge with a known RC constant and check discharge with something low burden as to not add undue discharge.

i bet this could be incorporated into the $20 LCR project, or at least use it as a base for such a thing.

[MiroS]

I played a bit with my tester , increased sensitivity to nA range and  I concluded that  you have to do test at operating voltage otherwise you will not detect cap failure, e.g. I had a 4.7uF/400V cup which is not leaking at all (zero nA) till 29V  and at 30V jumped rapidly to 250 nA. It seems to me that low voltage testers like MrCarslons or mine are usefull only for specific problems and are not giving ultimate answer.

Anyway I understand a magic of blinking  LEDs for electronics hobbyist , something like flute musik for snakes 

[Beamin]

Beaman, hes not interested in explaining how it works because he wants to monitise his activity on the internet.  Heres the business plan, first you hook em for free, then you stop providing free content, then you get em to pay.  There is nothing wrong with this its just a little anoying because he comes across as a Dudley Do Right, out there looking out for the common man.  As others have pointed out on this form, he seems a little creepy.

I think its significant that he hasnt released a video explaining how the fucking thing works.  I also find it interesting he is not responding here about this tester he is so cryptic about, except if you pay for his channel, and even then I not sure you will find out.

I subscribed to this patreon thanks to another member watched the video couldn't figure it out. He goes over it but every video I have watched, twice sometimes, I have understood things fully. So I left a post saying that I subscribed to figure out the mystery. But if that doesn't work I'll post a small part of the schematic blocking out the other parts(how the lights work etc.) of the circuit so people can't copy/ build it without the patreon. Then I will see if we can figure it out. It's only four transistors, no crystal or any  oscillator I could see. It seems like it amplifies the leakage and has this diode that had to be very stable and not exposed to heat while soldering. I think he's secretive because if people saw it they would say "all this hype for a couple of transistors? I could have built that (but they didn't and weren't first to think of it)". When ever a video is trying to sell something and has no dislikes or comments enabled I usually stay the fuck away. That maybe his undoing. If I was a 1st time viewer I would have just called bs and not subscribed.

[IanB]

You're looking at this the wrong way.

What someone reading this thread would do, is watch public videos showing the operation of the device, then sit down with pencil and paper and work out how to design something that behaves the same way.

You posted in another thread, "How would you go about designing something?" Well, here's your opportunity.

Start in this thread. Begin by describing exactly how the device behaves (that's your objective).

Then start proposing functional blocks that could reproduce that behavior. Get feedback from the forum if you get stuck.

Next start thinking about how to implement those functional blocks.

Last, build a prototype and test it. If (when) the prototype doesn't work, debug it, adjust it, and iterate until it does.

Do that and you will be way ahead.