Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives :: Books
Support This Site: Contributors :: Advertise


It is currently Jun Thu 21, 2018 3:10 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 7 posts ] 
Author Message
 Post subject: Capacitor leak detector
PostPosted: Mar Thu 08, 2018 10:42 pm 
Member
User avatar

Joined: Jan Wed 11, 2012 5:12 pm
Posts: 525
Location: Colorado Springs, CO
After a lot of study of capacitor leakage and insulation resistance (IR) I have come up with what I believe is a safe, simple, inexpensive, practical cap leak detector. The best part is that the heart of it, the DC-DC boost module is actually a freebie from Harbor Freight, an electric fly swatter.

I had done some experimenting with leak testing using a large number of 9v batteries, which worked, but not at rated voltage. I looked for DC-DC boost modules on line and couldn’t find anything with high enough voltage for a price I was willing to pay. Then I discovered that the fly swatter puts out about 1KV to an open load from two AA cells.

The output voltage is controlled with input voltage, however the input voltage was hard to control, so I added a voltage regulator, a LM2596S Buck Adjustable Power Supply module. Now a wide range of input voltages and varying loads would always deliver the desired output. I found this on ebay, six of them for $8 with free shipping. What a deal.
Attachment:
File comment: The schematic
Cap test.jpg
Cap test.jpg [ 45.56 KiB | Viewed 568 times ]

I am able to measure the very large IR indirectly by comparing the test voltage to the voltage drop across the DVM (10 megohm input resistance). Yeah, it’s really a DMM, I’m just using the voltmeter function for this. The best data I can find indicates that film caps less than .047mf should have IR of 10,000 megohms or greater. So, the IR of a good cap is about three orders of magnitude larger than the DVM. Therefore, a cap with good IR should register on the DVM about 1/1000 of what is dropped across the test cap. Larger voltage registered on the DVM indicates leakage/lowered IR. So, with a test voltage of say 630 volts, the DVM should register about .63 volts or less on a cap with good IR. For caps larger than .047mf it gets more complicated. Larger caps have lower IR and the capacitance must be taken into account when determining what leakage is allowable. I don’t know if testing electyrolytics is practical or what leakage is acceptable.

I added an external 10K precision ten turn pot in place of the on-board unit for ease of adjustment and a small meter to monitor the input voltage to the boost module, so as not to exceed the original 3 volt input. Due to the minimum output of the buck module being about 1.2 volts, I included a resistor and switch to give a high and low output. In the low range I am able to test from 100 to 200 volts, and in the high range from 300 to 750 volts. I then added a switch to short the input to the DVM to allow for faster charging and discharging of the test cap.

I don’t claim this is a cap checker, just a leak detector. In testing several dozen paper caps pulled from radios, and numerous new film caps, I have high confidence I can tell when a cap is leaky. I also found that if a cap shows leakage at lower test volts it never gets any better as the voltage is increased. If it leaks at 100 volts it will leak more at 630 volts. No need to test any further.

I see several positive aspects of this unit. It is safe, in that the main power source is batteries. So, no connection to AC power, which also makes it portable. The output of the boost module is sufficient to provide the very small current required, yet is easily pulled down if the current draw is too large. Could it be lethal? Maybe. It should be respected regardless. Also, with the DVM in series, a shorted cap will just show up as full test voltage on the DVM, no sparks, or smoke. I don’t know what happens if the meter short switch is closed while charging a shorted cap.

The reason I’m using a 1 megohm DVM (another Harbor Freight freebie) for the test voltage measurement is that its relatively low input resistance allows for quicker discharge of the cap when the shorting switch is closed.

The biggest problem I have found is that there is high frequency switching noise in the output of the boost module. This causes the DVM to “hunt” at the lowest voltage settings so that on a cap with low leakage the reading is unsteady. It has no effect on a leaky cap. I have tried using a choke in the output wire to no avail. This needs more study.

I installed the components in a cardboard box as proof of concept. I’ll do something better now that I know it works. Here are some photos.
Attachment:
File comment: Bottom side
2018-03-08_12-29-34.JPG
2018-03-08_12-29-34.JPG [ 198.37 KiB | Viewed 568 times ]

Leakage volts should be less than .10v. I know "leakage volts" isn't a real unit. If you want to know the IR you can do the math.
Attachment:
File comment: A .02 mf / 600v paper cap at 106 test volts, 18.3 volts leakage.
2018-03-08_12-36-35.JPG
2018-03-08_12-36-35.JPG [ 221.67 KiB | Viewed 568 times ]

Leakage volts should be less than .30v.
Attachment:
File comment: Same cap at 306 volts test, 75.4 volts leakage.
2018-03-08_12-37-01.JPG
2018-03-08_12-37-01.JPG [ 224.14 KiB | Viewed 568 times ]

Leakage volts should be less than .50v.


Attachments:
File comment: Same cap at 500 volts test, 148.1 volts leakage.
2018-03-08_12-37-25.JPG
2018-03-08_12-37-25.JPG [ 228.1 KiB | Viewed 568 times ]
File comment: New .022 mf / 630 volt film cap, 101 volts test, nil leakage.
2018-03-08_12-38-44.JPG
2018-03-08_12-38-44.JPG [ 231.29 KiB | Viewed 568 times ]
File comment: Same cap at 503 volts test, .069 volts leakage.
2018-03-08_12-39-25.JPG
2018-03-08_12-39-25.JPG [ 229.59 KiB | Viewed 568 times ]

_________________
I'd rather be on the ground wishing I was in the air,
than in the air wishing I was on the ground.
Top
 Profile  
 
 Post subject: Re: Capacitor leak detector
PostPosted: Mar Fri 09, 2018 4:04 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 513
Location: Equinunk PA 18417
I think your creative use of freebies is wonderful! I'm used to seeing leakage current expressed in microamps. You are making a similar measurement but expressing it as a voltage drop across something you are assuming is a resistor (1 meg input DMM).

Quote:
The biggest problem I have found is that there is high frequency switching noise in the output of the boost module. This causes the DVM to “hunt” at the lowest voltage settings so that on a cap with low leakage the reading is unsteady. It has no effect on a leaky cap. I have tried using a choke in the output wire to no avail. This needs more study.

Seems to me this compromises the integrity of whatever readout you get, doesn't it? The switching noise is AC and capacitors are exceedingly good at allowing high frequency AC signals through themselves. Your DMM is having trouble figuring out what it should measure. There are at least two possible sources of this noise. The HV boost converter is one. Rather than an inductor on the voltage output, I would be inclined to try a resistor and capacitor, as below. If it doesn't help the converter may be radiating noise in all directions and would benefit from some shielding.

My 10meg DMM generates a small AC signal that can be imparted to the circuit under test. I'll bet your 1 meg freebie does also. I think the following will give you nearly the same readings but with an advantage that both the HV supply and your cheapo DVM are isolated from the circuit under test (which unfortunately still includes the 10 meg DVM). Your freebie meter also serves to discharge the filter cap. A .01 ceramic disc would be a good choice, I think 60 cents from Mouser.

edit: schematic and text corrected


Attachments:
Cap testv2.jpg
Cap testv2.jpg [ 68.41 KiB | Viewed 423 times ]

_________________
-Richard Fairbanks
Top
 Profile  
 
 Post subject: Re: Capacitor leak detector
PostPosted: Mar Sat 10, 2018 5:15 am 
Member
User avatar

Joined: Jan Wed 11, 2012 5:12 pm
Posts: 525
Location: Colorado Springs, CO
Quote:
voltage drop across something you are assuming is a resistor (1 meg input DMM)

Yes, and no. It's across the 10 meg input DMM. The resistance could be expressed as ohms, or current in microamps, but it's easier to just use the voltage rather than calculating each measurement.
Quote:
Seems to me this compromises the integrity of whatever readout you get, doesn't it?

As far as I can tell it's only a problem when the noise is a significant portion of the readout, (S/N ratio). When a cap is leaky the measurement is relatively large and the reading is steady. It shows up only on caps with little leakage/millivolt readings. I have assumed that the noise comes from the boost module. I should put the o'scope on the buck module looking for noise. I like the idea of the resistor & cap to filter it out. There is already a .33mf cap across the output of the boost module. I was going to add them to the circuit this afternoon, but the boost module seems to have conked out. Maybe I've been pushing it too hard. There's not much to go wrong on it, a NPN transistor, transformer, diode, resistor and cap. I suspect the 5080 transistor died. If I can't figure it out I can wait for a flyswatter coupon, or just buy one for $2-3.
Quote:
Your freebie meter also serves to discharge the filter cap.

That's why a "better", high input resistance meter would not work as well.

_________________
I'd rather be on the ground wishing I was in the air,
than in the air wishing I was on the ground.


Top
 Profile  
 
 Post subject: Re: Capacitor leak detector
PostPosted: Mar Sat 10, 2018 6:26 am 
Member
User avatar

Joined: Jan Wed 11, 2012 5:12 pm
Posts: 525
Location: Colorado Springs, CO
I found that one leg of the transformer is open, so I'll have to get another module.

_________________
I'd rather be on the ground wishing I was in the air,
than in the air wishing I was on the ground.


Top
 Profile  
 
 Post subject: Re: Capacitor leak detector
PostPosted: Mar Sat 10, 2018 4:09 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 513
Location: Equinunk PA 18417
Too bad, no chance of finding one of those to replace it with.

_________________
-Richard Fairbanks


Top
 Profile  
 
 Post subject: Re: Capacitor leak detector
PostPosted: Mar Sat 10, 2018 4:52 pm 
Member
User avatar

Joined: Jan Wed 11, 2012 5:12 pm
Posts: 525
Location: Colorado Springs, CO
I'll just go buy a couple of them, they're cheap. I have a theory that the fast charging may overload it, especially on larger caps or shorted caps. Maybe some resistance there would be in order.

_________________
I'd rather be on the ground wishing I was in the air,
than in the air wishing I was on the ground.


Top
 Profile  
 
 Post subject: Re: Capacitor leak detector
PostPosted: Mar Sat 10, 2018 8:26 pm 
Member

Joined: Nov Wed 08, 2017 11:52 pm
Posts: 40
Location: Arizona
Why don't you get a more decent high voltage power supply, at least something like this https://www.ebay.com/itm/5V-12V-Step-up-to-300V-1200V-High-Voltage-DC-DC-Boost-Converter-Power-P3M2/112533508877?epid=2201866668&hash=item1a3385130d:g:AX4AAOSwhMpZmbi3.
These Fly Swatters are crap, they cannot deliver ANY constant load, they can barely slowly charge a capacitor. If you have a real leaky capacitor I am sure the output voltage will collapse.


Top
 Profile  
 
Post New Topic Post Reply  [ 7 posts ] 

All times are UTC [ DST ]


Who is online

Users browsing this forum: calmark, Indiana Radios and 7 guests



Search for:
Jump to:  
























Privacy Policy :: Powered by phpBB