Maximum voltage rating for a resistor

[advark]

Hi everyone,,

I was just read that post https://www.eevblog.com/forum/beginners/resistor-maximum-working-voltage/msg630976/#msg630976 and since English is not my native language I just want to make sure I understand this correctly before I blown myself up  .

Let say I put a 10M resistor, that is rated voltage of 250V,  on a 1KV rail, something will go wrong for sure. As I understand, since that resistor can handle up to 250V, if I used 4x 2.5M resistors, rated 250V, in series it will work since each of them will handle 25% of the total voltage drop, i.e. 250V each (to be safer, I should use 5x 2M resistors). Am I right?

Many thanks.

[retrolefty]

Yep, you got idea correct. 

[IanB]

There are two issues to be concerned with: (1) the power dissipation of the resistor and (2) the breakdown voltage or flash over voltage across the resistor.

Let's say you put 1 kV across a 10 M resistor. Then the power dissipation will be V²/R = 1000² / 10e6 = 0.1 W. So if this resistor were, say ¼ W it would be OK where heating is concerned.

To handle the breakdown voltage you should put several resistors in series, as you propose, so that the voltage per resistor is below the limiting threshold.

[coppice]

Yes, you are right, and it is common practice to do that. Beware, however, that you need to lay them out well to maintain their high voltage performance - e.g. no zig-zag layouts, which bring points of high voltage difference close together. Beware also that in some countries, in some applications, it is a requirement to use a single high voltage resistor for regulatory reasons.

[AF6LJ]

This gets into an aria of design that these days is not normally an issue.
I brought this up in the vacuum tube function generator thread as a concern when replacing resistors in old vacuum tube gear. This came up for me when I was rehabilitating a high power linear amplifier for amateur radio use. The metering circuit that was intended ot measure the amplifier's anode DC voltage used three 4.7 megaohm resistors in series with a lower value resistor in the series string to ground , The meter was placed across this low value resistor. That resistor was fine as far as the voltage drop across it went, even when the other three were exposed to 1200V each one of them went high in value, real high and the meter was reading about 40% of the actual voltage. Thankfully Digikey had proper replacement resistors, although they were rated at holding off 1200V, they were the best I could get for replacement unless I wanted to layout a new PC board. The thought did cross my mind.

When you dig into the data sheet on the resistors you are using you will find a maximum hold off voltage, and you have to watch those Dale power resistors while they can hold off quite a bit of voltage between terminals the breakdown voltage to the metal housing is lower.

[advark]

There are two issues to be concerned with: (1) the power dissipation of the resistor and (2) the breakdown voltage or flash over voltage across the resistor.

Let's say you put 1 kV across a 10 M resistor. Then the power dissipation will be V²/R = 1000² / 10e6 = 0.1 W. So if this resistor were, say ¼ W it would be OK where heating is concerned.

To handle the breakdown voltage you should put several resistors in series, as you propose, so that the voltage per resistor is below the limiting threshold.

I didn't mention it, but yes I had considered the power issue.

Yes, you are right, and it is common practice to do that. Beware, however, that you need to lay them out well to maintain their high voltage performance - e.g. no zig-zag layouts, which bring points of high voltage difference close together. Beware also that in some countries, in some applications, it is a requirement to use a single high voltage resistor for regulatory reasons.

Thanks for pointing those out. I didn't know about the zig-zag layout. Point taken. I'm just building a little gadget to test the voltage on old light fixture ballast that can go up to 30KV so at this point the regulation is the least of my concerns

Cheers

[singapol]

If you are playing with 30KV then spend good money on proper 10M resistors.Insurance sellers will
try to put this logic across by saying " How much is your life worth?". I would add further to read about
high voltage insulation and such although you see Mr.Nikola Tesla sitting in the room with mega voltages arcing across the room. To be knowlegeable is to be prepared and not ask why things happen to you after the fact. Yes we can all wing on a prayer but are never too sure. In electrical or electronics the term "over specification" is your friend.

[Wim_L]

Also be careful when assuming resistors will retain their value at high voltages. We tend to assume the current-voltage graph of a resistor is linear, but that isn't always true. Especially at high voltages, there can be a noticeable voltage coefficient.

[Cliff Matthews]

I can get 2W 750v Flame Proof Carbon Film Resistors for 25-cents each. http://www.justradios.com/resistors.html  So if I wanted a 100:1 scope probe, 14 in series would yield a working voltage of 10.5Kv. Is it a good idea to coat the whole assembly with HV auto igntion spray before insertion into some home-brew plastic probe tip?

[Performa01]

I can get 2W 750v Flame Proof Carbon Film Resistors for 25-cents each. http://www.justradios.com/resistors.html  So if I wanted a 100:1 scope probe, 14 in series would yield a working voltage of 10.5Kv.

Scope probe ... hmmm - what bandwidth do you think you'll be able to get with this construction?

I think it's much more realistic to make this a DMM probe (10 MOhm termination) instead of a scope probe (1 MOhm termination).

[AF6LJ]

I can get 2W 750v Flame Proof Carbon Film Resistors for 25-cents each. http://www.justradios.com/resistors.html  So if I wanted a 100:1 scope probe, 14 in series would yield a working voltage of 10.5Kv. Is it a good idea to coat the whole assembly with HV auto igntion spray before insertion into some home-brew plastic probe tip?

If you are looking for high bandwidth be mindful that not all carbon resistors are not created equal. Due to the ways in which some of these resistors are made there can be a higher than expected amount of inductance.

[Cliff Matthews]

@Performa01 - I considered it for both instruments, but yeah I'd need another 9meg in series for the scope (I'd have to work out some switch config )

@AF6LJ - Oh yeah, these over heat burning in that spiral pattern.. increasing inductance. So it's better to use more of the 400v 1/2watt carbon graphite powder types then?

[AF6LJ]

@Performa01 - I considered it for both instruments, but yeah I'd need another 9meg in series for the scope (I'd have to work out some switch config )

@AF6LJ - Oh yeah, these over heat burning in that spiral pattern.. increasing inductance. So it's better to use more of the 400v 1/2watt carbon graphite powder types then?

This is going to depend on the desired bandwidth, the inductance can be measured and can be found on some data sheets. Generally it is very small. If you are looking for a hundred MHZ bandwidth this could be a problem. A nano henry at a hundred MHZ is a significant amount of inductive reactince, twenty HMZ not so much...

[Performa01]

Where would you put another 9M in series? HV-probes are just one resistor string in series with the instrument. So for a scope, you need 99M total resistance, whereas for a DMM it would have to be 990M, considering it is to be a 100:1 probe.

As for the frequency response, you'll probably be very disappointed. All film resistors have a high inductance, except maybe very low values <100Ohm or so. So the first thing you want to check is that particular property in the datasheets - if specified at all, that is. Your best bet here would be carbon composite resistors.

But even if you have resistors with very low inductance (or even 'zero'), you'll still get an RC lowpass filter. Consider the 99M of the probe and say 20pF of the scope input - the resulting -3dB bandwidth will be some 80Hz at best. So to overcome this, you'd have to add a frequency compensation, which is far from being trivial for a HV-probe, all the more so if you cannot actually get inductance-free resistors in the first place.

[Cliff Matthews]

..for a scope, you need 99M total resistance, whereas for a DMM it would have to be 990M, considering it is to be a 100:1 probe.

Ok, got it for the DMM output. For my scope (mine is 18pf 1Meg), do I need any cap compensation for a max freq. of ~40Khz? There would be 5ft of coax with BNC's.

[T3sl4co1l]

Let say I put a 10M resistor, that is rated voltage of 250V,  on a 1KV rail, something will go wrong for sure.

Actually, "for sure" is unlikely -- they aren't going to break down at exactly the rating.  But you have no idea where they actually will, and the manufacturer provides this rating as a "it's pretty unlikely to break at this voltage", given the range of environment it's expected to be in (presumably, a certain amount of solder flux, humidity, dust, etc.).  (I don't know if they're actually certified to tolerate some combination of this, I'm just guessing.)

That said, it's worthwhile to use several more than the required (minimum) rating: it's very likely that, if one fails, it will fail more or less shorted, and increase the stress on others by as much.  So with 4 being rated minimum, 5 is not a bad idea, but I might even go with 6 or 8.  Or 9 or 10 just because the value will come out better (e.g., 9 x 1M into a 1M resistor, for a 10:1 voltage sense divider).

Amazingly enough, there are small (0603 to 1206 sized) chip resistors rated for extreme voltages, like 10kV (at least in peak).  The "gotcha" is, they have to be potted to reach that, as you might expect given the size of the parts!

I would recommend 1-2W through-hole resistors, maybe 3-4 in series (however many are needed for the rating, using the above method), though I don't think there's anything wrong with using a larger number of chip resistors in series.

By the way, if you're using this for AC measurement as well, even just for low frequencies, it may be necessary to add capacitors across the resistors.  Try to use one capacitor in parallel with each resistor.  This keeps the voltage equalized along the chain, and compensates for capacitance to ground (which varies along the chain, and makes weird dips in the frequency response).  You also need a capacitor across the 'bottom' resistor, and in a large ratio divider (like 100:1 or more), it needs to be quite large in comparison, and preferably, adjustable.

Tim

[AF6LJ]

A lot of good suggestions here, one other thing. use high voltage lead dress for most of that divider chain since the potential for corona is high.
Smooth round connections no shapr edges and you should be fine.

[Performa01]

..for a scope, you need 99M total resistance, whereas for a DMM it would have to be 990M, considering it is to be a 100:1 probe.

Ok, got it for the DMM output. For my scope (mine is 18pf 1Meg), do I need any cap compensation for a max freq. of ~40Khz? There would be 5ft of coax with BNC's.

Ok, according to the schematics your probe is going to be 100:1 for the scope and 10:1 for the DMM, that looks good to me.

The problem with the frequency compensation is that the resistors themselves will have some capacitance and I guess this would be even more of a problem than the inductance in your application. For proper compensation you want the reactances of your resistor string and the scope input to be the exact same ratio as the resistors, i.e. 100:1. If your scope has 18pF at the input, the resistor string should have 18pF/99 ~ 0.18pF in parallel, which is probably lower than the stray capacitance of the resistors itself. In other words, without any compensation it might well be that the output of your probe would not drop by 3dB at 80Hz (as I've stated in my previous post) but gives way too high an output voltage at higher frequencies.

For reliable results, you have to lower the impedance for ac, i.e. increase the capacitance values to a point where the stray capacitances become insignificant. You have 30 resistors in series, so you can put 100pF in parallel to each one, this will give a total capacitance of 3.33pF. In order to achieve a 100:1 ratio, you now need 330pF at the scope input. Subtract some 150pF for the coax cable and 18pF for the scope itself, there will be about 162pF left that need to be in parallel with the scope input. You would use just another 100pF here, in parallel with a 100pF trimmer cap. Now you should be able to adjust the probe compensation just as for any other 100:1 probe.

The capacitors must be rated for at least 1/30 of the total voltage, so if you plan to use your probe up to 10kV it would be 333V. 500V ceramic caps are cheap, and as NP0 (CG0) they also have good electrical properties. Alternatively you could also use polypropylene caps rated for 400 or 630V. The voltage rating of the trimmer cap is less critical as it is parallel to the scope input and will never see more than 100V in this scenario. Small (5mm diameter) ceramic trimmers usually have this voltage rating, but slightly bigger (10mm) foil trimmers will provide 250V.

If you build this, it would be nice if you'd show us the results, e.g. what a 1MHz squarewave looks like on the scope with your probe... 

[AF6LJ]

I would love to see how it works out as well.

[Cliff Matthews]

I would love to see how it works out as well.

Our rumination has been good, but components seem headed north of $50 and unfortunately, its usefulness would be sparse - other projects wait..

[AF6LJ]

I would love to see how it works out as well.

Our rumination has been good, but components seem headed north of $50 and unfortunately, its usefulness would be sparse - other projects wait..

You have me thinking about building one as an in-between major projects, project.
This has been a good exercise. 

[Cliff Matthews]

I would love to see how it works out as well.

Our rumination has been good, but components seem headed north of $50 and unfortunately, its usefulness would be sparse - other projects wait..

You have me thinking about building one as an in-between major projects, project.
This has been a good exercise.

Indeed, I searched earlier posts on HV potting compounds and found Allied has an MG chemicals 2-part kit big enough to pot 2 or 3 probes for $35

Other data sheets say it has a shelf life of >3 years (after opening?) http://www.alliedelec.com/mg-chemicals-832c-375ml/70125822/
Sure would be nice if Tim(T3sl4co1l) or Allan(W2AEW) were able to impart wisdom on the 100pf caps, trimmer, and probe plastics. It would be a pita, after assembly, if the chosen plastic put someone in the hospital (or worse..xx). 

[T3sl4co1l]

Sure would be nice if Tim(T3sl4co1l) or Allan(W2AEW) were able to impart wisdom on the 100pf caps, trimmer, and probe plastics. It would be a pita, after assembly, if the chosen plastic put someone in the hospital (or worse..xx).

Eh?

Oh, were we doing a thing now..?

*Inspects thread history*

Oh, I guess you kind of hijacked this thread, didn't you (after the OP's question was, I guess, well enough answered).  I don't know what other background you had in mind though, you didn't introduce a project or anything...

So, have you got a schematic, or description and specs, in mind..?

FYI, a DMM probe is a much more tricky thing to create.  You will want to check the resistance of your DMM very accurately: as accurate as your desired probe accuracy!  For instance, my DMM (an off brand, but pretty good performance) is something oddball like 10.8M and some change.  The exact value doesn't matter because it's intended for accurate measurements on lower impedance sources (it gets stated digits for sources under 1kohm or so -- who cares if it's exactly 10.000M in that case).  It would be very tricky to calibrate against if the DMM happens to be a bit above 10M, and very tricky indeed if it's a little below!  (If it's above, you can connect a resistor in parallel with the DMM to shunt the resistance down; but you need a very large resistance, like a 100M trimmer, to get the right range.)

Tim

[Cliff Matthews]

I'm truly sorry Tim  It just crept-up on me.. Like a corona crossing first one resistor, then two, then a whole string and.. well you get the picture. After watching Dave's video's on DMM false CAT ratings, I think I'd feel safer with a probe I can trust above 220v (with gloves, I'm like chicken little).