Have any of you considered making a high-pot tester for isolation-testing?
I would suspect that it "only" requires a voltage around 3-6Kv with very low power? Google shows many examples, but some of the implementation-details would require extreme care?
I would use it for testing smps-transformers and complete smps-modules.
Kind regards TroelsM
I would suspect that it "only" requires a voltage around 3-6Kv with very low power? Google shows many examples, but some of the implementation-details would require extreme care?
I would use it for testing smps-transformers and complete smps-modules.
Kind regards TroelsM
There is low possibility of a reliable outcome from home made HiPot testers.
This is an industry standard machine; FLUKE 1550C | Fluke 1550C, Insulation Tester, 5000V, 1TΩ, CAT III 1000V | RS Components
If just to play with at home, a lineoutput transformer from a monochrome Television will provide around 7kv how you will measure up to 1TΩ is anyones guess but it will be lethal to use without serious current limiting.
This is an industry standard machine; FLUKE 1550C | Fluke 1550C, Insulation Tester, 5000V, 1TΩ, CAT III 1000V | RS Components
If just to play with at home, a lineoutput transformer from a monochrome Television will provide around 7kv how you will measure up to 1TΩ is anyones guess but it will be lethal to use without serious current limiting.
Hi. I understand that a homemade variant will never be as trusty as a Fluke, - but as most amateurs will never get a Fluke, I guess that any testing is better than no testing?
Lets say that a homemade unit could provide between 1kv and 5Kv ( variable with a variac on the pri.side). With suitable resistors it should be possible to measure the HV with a moving coil-meter?
This doesn't have to be precise, to be useful. If my homemade transformer breaks down at 1Kv its not good enough, -regardless of whether the voltage was actually 750V or 1500V. If a transformer passes a test with approx 2Kv I would trust it a lot more than a transformer with no test at all.
It appears to me that the most valuable test would be a break-down test. Thus actual resistance-test is not needed?
Does that make sense?
Kind regards TroelsM
Lets say that a homemade unit could provide between 1kv and 5Kv ( variable with a variac on the pri.side). With suitable resistors it should be possible to measure the HV with a moving coil-meter?
This doesn't have to be precise, to be useful. If my homemade transformer breaks down at 1Kv its not good enough, -regardless of whether the voltage was actually 750V or 1500V. If a transformer passes a test with approx 2Kv I would trust it a lot more than a transformer with no test at all.
It appears to me that the most valuable test would be a break-down test. Thus actual resistance-test is not needed?
Does that make sense?
Kind regards TroelsM
Dear Jon, you must have deeper pockets than mine and assume troelsm has same. 😀
Dear Troelsm: you KNOW you must be VERY careful.
Link to some of those projects, the ones which look more reasonable.
In the old days, the grandfather of HiPot testers was the Insulation Tester, commonly called the Megger.
Megohms being a known concept, Teraohms not even in SF novels.
Original ones , used by Telegraph and Phone Companies since the 30´s at least, used the then ubiquitous hand cranked generator mounted in phones to call the Operator, plus some magneto type switches to get about 500V test voltage, current could be read on an analog meter and reading was calibrated in Megohms.
Just a typical analog multimeter type resistance meter, but with 500V DC instead of some battery voltage.
Then one day (think the 70´s) when I saw at one of those wonderful British Hobby Magazines (might have been ETI) the "Electronic Megger" project, basically a battery powered transistor oscillator into a ferrite core, getting 500V DC and reading current with a basic needle meter.
Basic same concept.
I bet a DIY one will work for your own use, specially when developing a new SMPS or winding a transformer, but remember that test carries zero weight outside your home, if you wan to certify a product you need a registered Lab to issue it.
Keep us updated, and post any interesting link you found.
In any case limit current to WAY less than 1mA, maybe 100uA tops, and do NOT approach your test subject with two crocodile clips carrying a couple kV between them, rather apply test probes first , with meter system turned OFF and UNPLUGGED, then from a distance plug and turn it ON.
Dear Troelsm: you KNOW you must be VERY careful.
Link to some of those projects, the ones which look more reasonable.
In the old days, the grandfather of HiPot testers was the Insulation Tester, commonly called the Megger.
Megohms being a known concept, Teraohms not even in SF novels.
Original ones , used by Telegraph and Phone Companies since the 30´s at least, used the then ubiquitous hand cranked generator mounted in phones to call the Operator, plus some magneto type switches to get about 500V test voltage, current could be read on an analog meter and reading was calibrated in Megohms.
Just a typical analog multimeter type resistance meter, but with 500V DC instead of some battery voltage.
Then one day (think the 70´s) when I saw at one of those wonderful British Hobby Magazines (might have been ETI) the "Electronic Megger" project, basically a battery powered transistor oscillator into a ferrite core, getting 500V DC and reading current with a basic needle meter.
Basic same concept.
I bet a DIY one will work for your own use, specially when developing a new SMPS or winding a transformer, but remember that test carries zero weight outside your home, if you wan to certify a product you need a registered Lab to issue it.
Keep us updated, and post any interesting link you found.
In any case limit current to WAY less than 1mA, maybe 100uA tops, and do NOT approach your test subject with two crocodile clips carrying a couple kV between them, rather apply test probes first , with meter system turned OFF and UNPLUGGED, then from a distance plug and turn it ON.
In the US, eBay has older hipots that don't cost a lot. That is your safest bet.
If you plan on using say a microwave over transformer, you need to be very careful due to the available current. Any mistake can be lethal. Feed the transformer with a Variac and use a HV resistor to limit the output current. You can sense leakage current using a resistor in the secondary return.
A break-down is normally destructive. In most insulation it leaves a carbon-track which greatly lowers the breakdown in later events.
Breakdown *may* make sense in prototyping or sample-testing when you have large quantities in production. I can't see the sense of it in one-off construction.
Agree.
Allowing for breakdown to happen makes sense in a production line setting, where you have many units and can sacrifice a few; for home/hobby use I would be very happy just to confirm: "It does NOT breakdown with 1500 or 3000V applied"
Or your testing will turn expensive very fast 😉
Allowing for breakdown to happen makes sense in a production line setting, where you have many units and can sacrifice a few; for home/hobby use I would be very happy just to confirm: "It does NOT breakdown with 1500 or 3000V applied"
Or your testing will turn expensive very fast 😉
Yes, I think we agree. I would not test every unit until breakdown. But testing a single prototype, or testing single components such as wire, tape, lacquer would make sense to me.
Testing every transformer to a "medium high voltage" around 1500V would probably be a good thing.
Yes, I'm fully aware that any diy test will be worthless outside my own home and private use.
Thanks for all the input so far
Kind regards TroelsM
This tester delivers a very small amount of energy to the sample, and multiple successive tests do not show a decrease in breakdown voltage, even for sensitive materials, like ordinary paper:
A Low-Lethality Dielectric Strength Tester
A Low-Lethality Dielectric Strength Tester
Apart from Elvee's diy method, two other options comes to mind.
Vintage industrial insulation testers crop up on ebay/craigs etc sometimes. I just checked out my ebayed AVO RM215-L/2 Breakdown, leakage and Ionisation tester from circa 1985 and it is in excellent condition, with 6kVac and 12kVdc capability. I does need an isolation transformer before it, to provide a clean new neutral to earth link, as it uses the mains earth as an output terminal. The good thing is that it comes with quality parts and HV connections and probes, and metering, and current limiting, so for $30 was imho the simplest path to take.
Another option is to connect a cheap modern handheld IR tester with 1kVdc output, in series with a slave HV dc supply (eg. 2kVdc with variac control), to provide eg. up to a 3kVdc test supply, with 2Gigohm meter and 1.4mA current limit (courtesy of the IR meter). That doesn't provide an AC supply as may be standard for some types of test. If you get a cheap second hand IR tester (mine was $19 on ebay) then damaging it is no big deal. Of course all energised testing should be done 'hands off', with suitable redundant metering/indication of what could remain charged.
Ciao, Tim
Vintage industrial insulation testers crop up on ebay/craigs etc sometimes. I just checked out my ebayed AVO RM215-L/2 Breakdown, leakage and Ionisation tester from circa 1985 and it is in excellent condition, with 6kVac and 12kVdc capability. I does need an isolation transformer before it, to provide a clean new neutral to earth link, as it uses the mains earth as an output terminal. The good thing is that it comes with quality parts and HV connections and probes, and metering, and current limiting, so for $30 was imho the simplest path to take.
Another option is to connect a cheap modern handheld IR tester with 1kVdc output, in series with a slave HV dc supply (eg. 2kVdc with variac control), to provide eg. up to a 3kVdc test supply, with 2Gigohm meter and 1.4mA current limit (courtesy of the IR meter). That doesn't provide an AC supply as may be standard for some types of test. If you get a cheap second hand IR tester (mine was $19 on ebay) then damaging it is no big deal. Of course all energised testing should be done 'hands off', with suitable redundant metering/indication of what could remain charged.
Ciao, Tim
Last edited:
> a cheap modern handheld IR tester
FWIW: that item is nearly unknown in the US, a small niche market. (Unlike UK and kin areas where an insulation test is required on building wiring.)
https://www.amazon.com/insulation-resistance-tester/s?k=insulation+resistance+tester
Aside from "modern", I can find a classic "MEGGER", either name-brand or the same idea made and labeled in China.
https://www.amazon.com/bobotron-Insulation-Megohm-Resistance-Megohmmeter/dp/B08KPY925S
Modern:
https://www.amazon.com/dp/B00O14LDKW/
https://www.amazon.com/dp/B06W58Q1QF/
> in series with a slave HV dc supply
Now I am wondering: since there seems to be some standard breakdown voltage which any field service tester should reach, do you really need a booster supply too?
FWIW: that item is nearly unknown in the US, a small niche market. (Unlike UK and kin areas where an insulation test is required on building wiring.)
https://www.amazon.com/insulation-resistance-tester/s?k=insulation+resistance+tester
Aside from "modern", I can find a classic "MEGGER", either name-brand or the same idea made and labeled in China.
https://www.amazon.com/bobotron-Insulation-Megohm-Resistance-Megohmmeter/dp/B08KPY925S
Modern:
https://www.amazon.com/dp/B00O14LDKW/
https://www.amazon.com/dp/B06W58Q1QF/
> in series with a slave HV dc supply
Now I am wondering: since there seems to be some standard breakdown voltage which any field service tester should reach, do you really need a booster supply too?
With respect to using a handheld IR tester in series with a slave HVdc supply, so as to increase the applied dc test voltage to a DUT, I have used this for valve diode PIV leakage testing on a batch of 5U4. I had initially tested the batch of diodes at the 1kVdc level of the IR tester, and then included the slave HV supply to increase the test level to 1.55kVdc.
I managed any breakdown risk by paralleling a 3kV diode across the IR tester terminals, such that the IR tester output could not be driven negative, and the IR tester itself has over-current limiting that pulls its output supply voltage down. As the IR tester was battery powered and handheld, it was manually configured first and was then untouched during the subsequent energisation of the slave HV supply (which itself had a parallel reverse 3kV diode across its output terminals), which had one output terminal earthed.
Yes in oz there has been a legacy verification requirement for IR testing of new installation of mains AC cables/distribution circuits, as well as AC mains circuitry in equipment under manufacture, which has allowed use of a 500V or 1kV IR test, and hence a reasonable number of secondhand testers pop up on ebay.
https://www.dalmura.com.au/static/QM-1492.jpg
I managed any breakdown risk by paralleling a 3kV diode across the IR tester terminals, such that the IR tester output could not be driven negative, and the IR tester itself has over-current limiting that pulls its output supply voltage down. As the IR tester was battery powered and handheld, it was manually configured first and was then untouched during the subsequent energisation of the slave HV supply (which itself had a parallel reverse 3kV diode across its output terminals), which had one output terminal earthed.
Yes in oz there has been a legacy verification requirement for IR testing of new installation of mains AC cables/distribution circuits, as well as AC mains circuitry in equipment under manufacture, which has allowed use of a 500V or 1kV IR test, and hence a reasonable number of secondhand testers pop up on ebay.
https://www.dalmura.com.au/static/QM-1492.jpg
Hi
I found an old "megger" in the stash. Probably older than me, but appears to be fully functional and generates around 600V when doing high-res measurements.
But, I would really like to test at 1-5kv and preferably with a very low energy to keep the testing non-destructive.
Elvee´s construction appears to be perfect for the task, but also a little complicated.
I stumbled across the link below for a simpler "break-down tester". I think the cap in the output should be smaller to limit energy, but otherwise it appears well throughout.
The transformer could probably be hand-wound?
High Voltage Circuits
Furthermore I found a small transformer on ebay that looks good. With the correct drive-circuit and HV output rectifier it could work. - maybe it would work with Elvee´s HV-section?
20KV high frequency high voltage transformer ignition coil inverter driver bN_HH | eBay
Kind regards TroelsM
I found an old "megger" in the stash. Probably older than me, but appears to be fully functional and generates around 600V when doing high-res measurements.
But, I would really like to test at 1-5kv and preferably with a very low energy to keep the testing non-destructive.
Elvee´s construction appears to be perfect for the task, but also a little complicated.
I stumbled across the link below for a simpler "break-down tester". I think the cap in the output should be smaller to limit energy, but otherwise it appears well throughout.
The transformer could probably be hand-wound?
High Voltage Circuits
Furthermore I found a small transformer on ebay that looks good. With the correct drive-circuit and HV output rectifier it could work. - maybe it would work with Elvee´s HV-section?
20KV high frequency high voltage transformer ignition coil inverter driver bN_HH | eBay
Kind regards TroelsM
If you have the time and interest to pursue then perhaps just go for it and report back.
You haven't elaborated on the parts to be tested and whether you have decided on test levels that align with some existing testing scheme (eg. from a standard) or not, so I doubt there is much more to advise unless you observe strange behaviour from a tester that you are preparing, or parts you are testing.
You haven't elaborated on the parts to be tested and whether you have decided on test levels that align with some existing testing scheme (eg. from a standard) or not, so I doubt there is much more to advise unless you observe strange behaviour from a tester that you are preparing, or parts you are testing.