Insulation Resistance Meter for repair and troubleshoot uses?

[e0ne199]

anyway how much output current does insulation tester produce? for example when i want to measure continuity between a 20kV-100kVA transformer primary phase windings by injecting voltage around 5kV..

I would not test winding continuity with a high voltage tester. That would be carried out with a micro-ohm meter for better accuracy, i.e. low voltage-high current, that would be more likely to find a bad connection or inter-turn fault than an insulation tester would.

The top brand 5kV testers such as Megger, Sonel, Chauvin Arnoux can supply around 5 to 8mA before the output voltage starts to drop off. Other manufacturer's testers tend to be around the 1.5 to 2mA mark similar to the 1000V insulation testers.

Kind regards

anyway is it possible that doing that continuity test will break the transformer itself? although it is a 24kV transformer..

[vinlove]

I see, but non the less I still would not personally use it. Have you ever considered that the readings that you are seeing when using are not actually the insulation breaking down of the cables as you seem to be thinking, but the insulation within some capacitors in the circuit? That is where my money would be gambled on.

Have you got a capacitor tester? If so, which one and what does it actually measure. When you say that it over loads the current on transmit, that to mind is pointing directly at capacitor issues and if you have one or more capacitors on a supply rail shorting out to ground, which is where they are mainly used on power rails to provide a clean and steady supply to the circuit, then the rail is going to try to supply the power required to transmit PLUS the leakage current to ground via the capacitor(s).

Most problems with electronics are capacitors, they dry out with age and if they are installed near to a heat source then this ageing and drying out process is speeded up.

Also remember that when you use a ohms tester or a insulation tester at various points in a circuit, without disconnecting the component leg from the circuit (not the ground) that you are going to get false readings because you are trying to test multiple items at once which you cannot do.

A 500v insulation tester is unlikely to damage capacitors used in a tube transmitter, as it cannot produce enough current to do so.
Unless there are low voltage solid state devices as well as tubes, I would say the chance of damage is minimal.

That said, you are correct, in that if the caps are showing leakage, this may be the problem.

The OP hasn't given quite enough information:-

"Over current"?

Overall high current from the Mains supply or whatever?
High current from the HT supply?
Excessive plate current in the PA tube?

Does the transmitter have a CW mode?
If so, when keyed in that mode, does it draw normal current?

In AM mode, does it draw excessive current when in transmit with no audio input?

Depending on the design, the  modulator may not only be out of circuit in the CW mode (if it exists), but powered down completely.
In that case, the overall current draw will increase in AM mode,--- as an old unit, it probably uses high level modulation.

In a perfect world this shouldn't happen, but try re-dipping the PA anode current, just in case.

It is the vintage Heathkit DX40-U.
When it is powered, and on to TUNE position, all seems fine. All the tubes lit up, and voltages OK from PSU, and the smoothing caps.

But the moment that it is set to AM mode, the 5U4G rectifier tube arched and sparked with flashing lights and frying noise.
Then when in most speedy way possible, powered the set off, and checked the 5U4G, the tube is totally dead - no continuity on the heater pins.

It is same in CW mode too.  I have replaced all the eletrolyc caps, and even made up silicon rectifier to save the 5Us.
5Us tube are too expensive to blow off every time the set was powered on for testing.

The silicon diode rectifier worked OK, but the output rectified dc voltages were a little higher than the 5U4G.

But it too was fried when the FUNCTION wafer switch set to AM mode. The heat resistors and diodes were OK, but the plastic base off the dead 5U4G I used to make up the diode rectifier were melting away with toxic smell and fireworks.

So I have been working on troubleshooting what is causing the over current killing the 5U4G rectifier, and now melting away the silicon diode rectifier plastic base.
I have been looking for some area where its likely causing the shorts in the circuit and  the old wires in the set causing possible shorts with the chassis. But still no joy.
In the process, I noticed, the sets some wiring had been altered by the previous owner on one area, but it looks not much significant causing the terrible over current. But I might take the extra electrolytic cap out and change the wiring to the original as in the schematic.

Maybe I must start a new thread for this topic. So, I was looking at an insulation resistance meter which might point out any insulation problems in the set.
Not tried yet, and not sure if this will work.

[jpb]

The one use I've made of one is to do a sanity check on ESD mats where the resistance is too high to register on a standard meter but does on a 1000V insulation meter.
It also is a way of using the 1000V scale on a DVM. Not very scientific - qualatitive rather than quantative though I do compare with some test multi-gigaohm resistors.

[vk6zgo]

I see, but non the less I still would not personally use it. Have you ever considered that the readings that you are seeing when using are not actually the insulation breaking down of the cables as you seem to be thinking, but the insulation within some capacitors in the circuit? That is where my money would be gambled on.

Have you got a capacitor tester? If so, which one and what does it actually measure. When you say that it over loads the current on transmit, that to mind is pointing directly at capacitor issues and if you have one or more capacitors on a supply rail shorting out to ground, which is where they are mainly used on power rails to provide a clean and steady supply to the circuit, then the rail is going to try to supply the power required to transmit PLUS the leakage current to ground via the capacitor(s).

Most problems with electronics are capacitors, they dry out with age and if they are installed near to a heat source then this ageing and drying out process is speeded up.

Also remember that when you use a ohms tester or a insulation tester at various points in a circuit, without disconnecting the component leg from the circuit (not the ground) that you are going to get false readings because you are trying to test multiple items at once which you cannot do.

A 500v insulation tester is unlikely to damage capacitors used in a tube transmitter, as it cannot produce enough current to do so.
Unless there are low voltage solid state devices as well as tubes, I would say the chance of damage is minimal.

That said, you are correct, in that if the caps are showing leakage, this may be the problem.

The OP hasn't given quite enough information:-

"Over current"?

Overall high current from the Mains supply or whatever?
High current from the HT supply?
Excessive plate current in the PA tube?

Does the transmitter have a CW mode?
If so, when keyed in that mode, does it draw normal current?

In AM mode, does it draw excessive current when in transmit with no audio input?

Depending on the design, the  modulator may not only be out of circuit in the CW mode (if it exists), but powered down completely.
In that case, the overall current draw will increase in AM mode,--- as an old unit, it probably uses high level modulation.

In a perfect world this shouldn't happen, but try re-dipping the PA anode current, just in case.

It is the vintage Heathkit DX40-U.
When it is powered, and on to TUNE position, all seems fine. All the tubes lit up, and voltages OK from PSU, and the smoothing caps.

But the moment that it is set to AM mode, the 5U4G rectifier tube arched and sparked with flashing lights and frying noise.
Then when in most speedy way possible, powered the set off, and checked the 5U4G, the tube is totally dead - no continuity on the heater pins.

It is same in CW mode too.  I have replaced all the eletrolyc caps, and even made up silicon rectifier to save the 5Us.
5Us tube are too expensive to blow off every time the set was powered on for testing.

The silicon diode rectifier worked OK, but the output rectified dc voltages were a little higher than the 5U4G.

But it too was fried when the FUNCTION wafer switch set to AM mode. The heat resistors and diodes were OK, but the plastic base off the dead 5U4G I used to make up the diode rectifier were melting away with toxic smell and fireworks.

So I have been working on troubleshooting what is causing the over current killing the 5U4G rectifier, and now melting away the silicon diode rectifier plastic base.
I have been looking for some area where its likely causing the shorts in the circuit and  the old wires in the set causing possible shorts with the chassis. But still no joy.
In the process, I noticed, the sets some wiring had been altered by the previous owner on one area, but it looks not much significant causing the terrible over current. But I might take the extra electrolytic cap out and change the wiring to the original as in the schematic.

Maybe I must start a new thread for this topic. So, I was looking at an insulation resistance meter which might point out any insulation problems in the set.
Not tried yet, and not sure if this will work.

Here are a couple of links which may be useful



http://www.w4btx.com/DX-40%20Manual.pdf

Looking at the manual in the second link, It may be worth checking  C33.
Note that the video in the first link had a lot of zapping in the mode switch, & the author repaired the switch, so it may be another possibility.

One trick is to actually measure the HT current draw, by breaking the HT feed &  inserting a multimeter on the amps range.

Be careful, because, when things are going OK, the HT is around 750v, which may be in excess of the voltage ratings of some cheaper DMMs.

[vinlove]

Thank you for the link. The video is excellent.

Will check out the C33. Last night I was checking out a few dodgy looking caps in the set, and found out that one of the cap was supposed to be 4700pF in the schema, but the actual cap was 47pF.

Could it be significant for causing the over current?  I will look into my junk box, and get some 4700pF ceramic caps out to replace the 47pF one.  Not sure which cap in the schema it is, it was too late to look into the schematic last night.

Yeah, most of my DMMs (I have a few cheap DMMs) says it can do upto 1000 DCV and 750 ACV.  My old Fluke 25 did read one day 1500 ACV.  But I get a bit nervous when trying to read HT.

[vinlove]

It is C46,   0.005uF = 5000pF 1.6kV in the schematic of the DX40 US version.
In my set, the cap was rated at 47pF 400V. That looks like grossly wrong rated cap, the previous owner put in.

Would it have contributed the serious over current fault in the set?

[Specmaster]

It is C46,   0.005uF = 5000pF 1.6kV in the schematic of the DX40 US version.
In my set, the cap was rated at 47pF 400V. That looks like grossly wrong rated cap, the previous owner put in.

Would it have contributed the serious over current fault in the set?

I did ask before but don't think I ever got an answer, what are you testing the capacitors with and what measurements are taking? If you are only using a meter that can read the capacity then that is not sufficient, yes it will show up any significant drift in value which may or may not be linked to your problem. You need to do a check on their ESR values using a meter capable of reading ESR and you also need to be able to check the capacitor for leakage and that involves testing with a far higher voltage then and ESR meter or a DMM meter is capable of supplying. You need to use something like a Heathkit C-3U bridge to test for leakage.

C46 could well be a smoking gun as with that amount of drift and voltage, it is more than likely to be leaking and could be the cause of your problem. Replace it and see. You could try, after discharging it, reading its resistance using a normal DMM and see if it is reading anything, if so replace it.

[vinlove]

It is C46,   0.005uF = 5000pF 1.6kV in the schematic of the DX40 US version.
In my set, the cap was rated at 47pF 400V. That looks like grossly wrong rated cap, the previous owner put in.

Would it have contributed the serious over current fault in the set?

I did ask before but don't think I ever got an answer, what are you testing the capacitors with and what measurements are taking? If you are only using a meter that can read the capacity then that is not sufficient, yes it will show up any significant drift in value which may or may not be linked to your problem. You need to do a check on their ESR values using a meter capable of reading ESR and you also need to be able to check the capacitor for leakage and that involves testing with a far higher voltage then and ESR meter or a DMM meter is capable of supplying. You need to use something like a Heathkit C-3U bridge to test for leakage.

C46 could well be a smoking gun as with that amount of drift and voltage, it is more than likely to be leaking and could be the cause of your problem. Replace it and see. You could try, after discharging it, reading its resistance using a normal DMM and see if it is reading anything, if so replace it.

Ahh.. sorry I missed that question I think.  I quickly drop by to the forum, and scanned msgs, so tend to miss some posts and questions at times.

Yeah, I do have an ESR reading meter which is BK-328, it displays ESR value. The fact is that the C46 is a new capacitor put in by the previous owner, but somehow he put the wrong valued one, because I can see the cap being very recent modern one, only replaced a few months ago?

Anyhow, I cannot tell now if it is directly linked to the problem of the over current without powering it on and testing it.  I was looking into my junk box, but couldn't find any good matching value of the cap for the replacement.  So I had to order some caps from eBay, and will take a few days to arrive, and I will wire the new cap on, and will power on for testing.

I am not sure if it is worth paying for the dedicated cap testing device like the vintage Heathkit. The problem is that they tend to be grossly over-priced for collectors value rather than what they do or how well they do it.

And it is quite impractical to desolder all the caps from the set, and go through the leakage testing when troubleshooting problems.  You just try to identify the problematic areas, and pick out suspicious components and do testing, and if in doubt or problematic for sure, then replace them, that is my strategy.  And capacitors are usually very cheap, so it is easy to replace them, if they look old or have usual suspicious tell tale signs for being bad.

And I am sure that reason that you are reading the capacitance is that when the value has drifted a lot from the nominal value, that would be the first sign that the cap is not good anymore, so it is vital to get the capacitance readings I thought.

All those old paper wax caps, when you take off from the set, and read the capacitance, they are usually well off from the nominal value, which tells a lot.

But really, this C46 was a very tricky cap, because it looked shiny new, and I took it for granted as a good cap, which turned out to be well out of value, not because it gone bad, but because it was simply a wrong cap for the junction in the circuit. If I have not tested it just purely by chance trying out my newly arrived DMM,  looking at the schematic, it would have never been noticed forever.

[Specmaster]

The actual capacitance, even if it is very close to the designated value can still be a false indication as I have found many caps that are within spec on their value but fail the leakage test and or the ESR test and don't forget that the ESR reading is also frequency dependant can pass testing at 100Hz but fail at 10KHz or at 1MHz etc. Like wise a cap rated at say 500VDC can pass a leakage test with ease at 16V, but fail as it gets closer to its rated V rating.

Clearly from your description it is something that is only switched into a live circuit when the set is switched to AM.

I think that if the 5U4GB is pulled from the circuit then you only have 5 pins to check for any signs of leakage to the chassis and with a DMM set to resistance mode then it should be possible to see if anything is leaking to ground.

Edit:
I take it that you have either tested or had the tubes tested as it could be possible that a tube is faulty and causing your problems as well?

[vinlove]

I found that capacitors testing is tricky issue, with all the multiple factors involved, as you pointed out.

I still will read capacitance first then ESR, when testing caps, but if it looks old or physically not sound, then just replace them with the same value and voltage rating. That is the way I do it.  For that, I would still keep the capacitance meter and ESR meter which were cheap, but I aint gonna pay over hundreds for the vintage Heathkit capacitor tester.

Yes, I made sure the 5U tubes were all sound. In fact the set has destroyed 3x new 5Us already in the middle of test powering on session. The moment it was switched to AM mode, the 5U tubes got fried instantly. And then I made up a few diode rectifiers to work as 5Us. They worked well, but the over current problem is still there. Hopefully the new cap replacement will make some difference to the situation.

[Specmaster]

I agree, keep all the capacitor testing you have but where in blue blazes did you get the idea that a Heathkit C-3U would cost hundreds from?? You don't state your country in your profile, which actually makes it harder for other members to assist you especially when giving you links to refer to which are often going to be country specific, so maybe you might like to revisit your profile page and rectify that issue for your own benefit.

I say Heathkit C-3U because that is the UK variant of that tester, if you are in the USA for instance than that reference becomes just a C-3.

Anyway, I can assure that your guess at the cost of that particular tester is way off, I got mine on eBay just a few weeks ago for only £34.95.

That Video in the thread is worth watching. It really is a simple piece of kit and it wouldn't cost a fortune to replace all the caps and resistors in and that would certainly eliminate those from the equation.

[trobbins]

Vinlove, have you restored vintage valve gear before?

I recommend you start with a schematic, and testing transformers for insulation resistance, winding resistance and winding turns ratio (to confirm no internal shorts), then rewire all insulated wires where the vintage insulation is suspect. Then test subsections by themselves, and use a variac and AC current meter, staring with just the AC mains side wiring and only primary winding connected to confirm acceptable idle current.  Any HV secondary side valve rectifiers should have redundant ss diodes inserted in series, and secondary windings should have appropriate fusing added where practical.  Then it is common to test with B+ but without valves, using a variac, to confirm no secondary side circuitry faults, and to allow coupling capacitor leakage to be tested, as well as testing B+ filter capacitors up to their max rated operating levels.

If you dive in to testing by powering up straightaway then you run the risk of damaging parts (as you have experienced), and sometimes those damaged parts can be irreplaceable, and you may end up being frustrated and still not able to localise faults.

[Specmaster]

Vinlove, have you restored vintage valve gear before?

I recommend you start with a schematic, and testing transformers for insulation resistance, winding resistance and winding turns ratio (to confirm no internal shorts), then rewire all insulated wires where the vintage insulation is suspect. Then test subsections by themselves, and use a variac and AC current meter, staring with just the AC mains side wiring and only primary winding connected to confirm acceptable idle current.  Any HV secondary side valve rectifiers should have redundant ss diodes inserted in series, and secondary windings should have appropriate fusing added where practical.  Then it is common to test with B+ but without valves, using a variac, to confirm no secondary side circuitry faults, and to allow coupling capacitor leakage to be tested, as well as testing B+ filter capacitors up to their max rated operating levels.

If you dive in to testing by powering up straightaway then you run the risk of damaging parts (as you have experienced), and sometimes those damaged parts can be irreplaceable, and you may end up being frustrated and still not able to localise faults.

In this instance, I think that we can ignore the testing of transformers seeing as the problem apparently only manifests itself when the unit is in AM mode.

[trobbins]

I suggest that nothing can be ignored as the equipment is vintage valve gear, and as the OP has not identified the likely vintage or provided photos then it could be significantly old and had a tough life, or been left in damp storage for decades, or ... the list goes on.  Imho it would take some confidence building to see it as just a likely single fault within an otherwise clean, looked after, recently serviced piece of valve gear.

[Specmaster]

The reason why I think the transformers can be skipped is that these are in circuit all the time and the OP has already stated the fault does not show until he switches the set into AM mode. The transformers IIRC are switched on via 1 switch only so as long as the set is on, the secondary are connected, may not all be under load though but if there was shorted turns, dampness etc then I'd have expected any such problems to exhibit it self regardless of what mode it was in.

[vinlove]

specmaster, I am in UK, and I was bidding for a Heathkit Capacitor tester recently, and my max bid was 38£, but the final bidding was well over 125 or something like that. If you bought it for around 35 then you have been lucky, and got the bargain. Good for you. I think Heathkit devices work, but there is definitely collectors value gets added to them when they are being sold on auctions.

But yeah, transformers are OK. If the mains transformer is bad, then you wouldn't get good HV, and really the set is only good for parting out I would say. The set had been powered under the lamp current limiter and also via Variac, so it was well protected for the over current. And indeed all looks fine until it is switched onto the AM mode, and bypassing the lamp current limiter and under full mains power, then it goes into fireworks. So, I was suspecting some wire or component in the set and chassis shorts somewhere?

But then also found out the C46 cap is wrong value, so will replace it and power on test again.  I was thinking of injecting 500V from the Insulation Resistance meter, and see how good the insulation on the set is with the old wires, but it is still just an idea.  But it stands as the last option for the testing.

[Specmaster]

Totally agree with the transformers and with the dim-bulb / variac approach but still disagree with the 500V insulation tester. If you watched that video, there was nothing in that video that would not have shown up just the same using a normal multimeter connected between chassis ground and the various points that you wanted to use the 500V tester on. If you do your ohms law calculations then 1 megohm leakage to ground is only going to result in a current flow of 0.00075A and resistance of 1 kilohm will result in a fault current of 0.75A and both values are well with the scope of even the cheapest of multimeters without risking possible damage. Incidently, referring to a comment you made earlier about de soldering capacitors, you only need de solder on end of them, not both the measure their capacity and ESR. As a general rule of thumb, never measure the value of any part with it still connected to the circuit because you will be measuring the values of anything else that has a current path through it at the same time and that might be possibly why the capacitor you measured 47pf. Did you test it while it was still connected at both ends to the rest of the circuit?

As to the Heathkit capacitor tester, I got mine in an eBay auction, here in the UK, you just have to be lucky and if you're patient enough, you'll pick one up in the end for a good price, just set up a standard search and save it on eBay and every time one is listed, you'll get an email informing you. I got mine on the 19th Jan this year using that method.

[vinlove]

Yeah, quite see your point about the 500V DC input insulation test. I was really being suspicious about the old wiring in the set, and some of them have deep burn scars with hot soldering iron.

 use the insulation resistance meter for usually checking long cable or long wire for its continuity and insulation. Sometimes I test heater and demonic appliances for their insulation, but wondered if it could be used for troubleshooting vintage tube  electronics sets too. But maybe it is not good idea from this thread opinions

Yeah, I took the C46 cap off from the set, and tested it.  I know that you never test caps while they are in the circuit for the reason you explained.  It is a bit of hassle desoldering it back after the testing, but thats just life.

I also have a LSR tester device called BK-328, and it can test caps, coils and transistors and resistors. On cap testing it reads capacitance but also ESR and % of leak. I find it very consistent and accurate. But not sure what ESR value and Leakage value is regarded as good or bad.

But yeah, if I can see the Heathkit cap tester for cheap, then I would buy it. But definitely will never pay over £30. Because I would never use it that often. Usually I find new caps are almost always fine. It is the old wax and paper caps are problem, and you can tell them by their looks, don't have to test them

[Specmaster]

When say new caps are generally fine, are you referring to electrolytic caps? If yes, then again, I disagree, many can be easily spotted by their appearance, but equally I have had to replace caps that looked to perfect externally because their readings had changed massively, often caused by using low temperature ones near to heat sources such as heat syncs, transformers etc, or just generally where the equipment has been used in a high ambient temperature etc.

Generally you can spot ones that you want to check first and these are ones that are directly connected to the power rails and ground, if these are leaky they will cause (depending on the amount of leakage) cause over current to flow and in severe cases will cause the power rail to fall significantly below its voltage or collapse completely in the case of a short circuit in a cap.

Paper and wax caps really should be replaced on sight, regardless of their appearance, they cannot be trusted, ceramic, micra and polyester caps are generally OK.