Enceladus wrote ...

klugesmith,

Granted I have never tried this exact application with 1/4W resistors, but I have seen them used in high voltage applications such as connected across 1n4007 diode strings to ensure balanced voltage drops across each diode while they are blocking, as well as in small Marx generators and bleeders for TC primary cap banks. You can coddle your resistors if you like but I have never had any problems with 1/4W resistors at up to 1kV as long as I haven't exceed the power rating. Besides, they're like 5 cents. Who cares?

Also, I am well aware of how component tolerances work, but I was just trying to outline how to do it fast and dirty without it pouring smoke the second it's powered on. With 10X 1M there should be ample time to get a reading before it overheats.

It's a contradiction to label a 1M resistor as being rated for 1/4W and 250V. Ohm's law is pretty inflexible that way. It's got to be one or the other. It makes no sense to explicitly display one rating
while secretly limiting the true rating to less than half of what it's supposed to be. If a 1M 1/4W resistor is not rated to dissipate any more than 1/16W, then it's not actually a 1/4W resistor at all, is it?

I honestly didn't even know there was such thing as carbon film resistor datasheets. They have no numbers or letters or insignia except the bands. How do you reference that?

I was thinking that a capacitive divider might also work too if power dissipation is that big of an issue (although I remain skeptical that it would be.) High voltage caps are usually a common item in high voltage hobbyist junk drawers. At 60Hz it's not at all difficult to get all the reactance you'd need to limit the current to a reasonable level with just 2 caps. Maybe something like a 200pF and a 1.8nF with the output taken across the 1.8nF. (Just a rough guesstimate)

Connecting several meters in series just seems ridiculous to me but it looks like thats what he wants to do. Good luck with that Scott. Hopefully you don't end up with a pile of dead meters. If you change your mind about building a divider Scott, you can PM me if you need help.

-J

klugesmith wrote ...

Go for it! We await your results.

Note 1. Looks like you chose R values to get an 11:1 voltage division. So we are both in the club of people who don't mind doing arithmetic to convert instrument readings to real values of the variable of interest.

Have you considered that 11:1 is the _unloaded_ division ratio, and that connecting your DMM will significantly change the ratio? The specs you pointed me to don't say whether the meter's input impedance on voltage ranges is 10 megohms or 1 megohm, and the difference is huge in this application.
You can find out by measuring 120 VAC with and without the voltage divider -- a very simple way to calibrate the divider.
Or by measuring a 9 volt battery directly, and with 1.2 meg in series with meter, and with 12 meg in series with the meter.

Voltage divider loading by the meter is also a consideration for capacitive dividers. With the C values given most recently, connecting a 1 megohm voltmeter (not uncommon at the low-cost end of DMM market) would make a high-pass filter with corner frequency of 80 Hz. You could still easily get accurate readings using the AC calibration trick given above. Waveform distortion probably isn't a big concern for you.

In either case, if you changed your low-side R or C to get division of approximately 100:1, the loading by meter would be that much less significant.

Note 2. I would skip the electrical tape. Arc-over is not the failure mode for the R's we're talking about, at these voltages. High temperatures and/or high electric fields, in the resistive layer and its junctions, lead to long-term resistance changes and increased failure rates.

Wrapping the 12M resistor with thermally-insulating tape will make it get hotter than it needs to be, if it's connected for a long time. That would call for derating the component power. Maybe not by much if the wire leads are left long, as a heat sink. Hypothetically, if that resistor were immersed in dielectric oil there'd be huge margins against arc-over and over-temperature. Then you could see whether 2,500 volts applied for a day, or for a year, makes the resistor change measurably. I bet you have no lack of MOT's!