Tweet
I'll try to remember what I was saying. 
DMMs are handy, but just like an analog meter, you have to know what's inside them, because when you attach them to the circuit, the meter becomes part of the circuit. First off, analog meters were pretty up-front about it - they rated their impedance in so-many ohms per volt. 20K ohms per volt was a normal reading. What that meant was that if you set a 20K 0hms-per-volt meter to 100V, the input resistance was 100x20K, or 2M. This worked out because the analog meter movement was fundamentally a current measuring movement, and everything was scaled from that current-full-scale sensitivity and the movement's resistance. So when you moved the scale to a higher voltage, it necessarily put more resistance in series.
DMMs **may** do this, they may not. They may use more "innovative" means of scaling. The DMM is fundamentally an A-D converter measuring the voltage across some external resistance. They may scale the input with a resistor divider, they may scale it with other means that do not result in an ohms-per-volt, and many I've seen simply quote an input resistance, period. That's usually 1M or 10M, and may or may not, probably not, change with scale setting. You'd have to dig into the schematic of the meter to know, and simple no-name meters may be almost anything.
Testing a coupling cap for leakage, you're looking for the cap to be an open circuit compared to perhaps several megs (i.e., the old 2M-10M of the analog meter setup). With a DMM of unknown input resistance, the measurement gets tricky. It would be nice if you know that the meter is 10M input resistance on all scales. Then you could simply hook the meter between the coupling cap "ground" side and ground, and look at voltage. The ratio of the voltage on the other side of the cap to the voltage you read is the number you multiply the meter resistance by to estimate the capacitor's leakage resistance.
You'd like the voltage you read to be less than 1% or so of the voltage on the high side of the cap, so that if you had a 10M DMM, the resistance would be no smaller than 100 times 10M, or a giga-ohm, which is sufficiently open circuit to make the AIR around the cap part of the leakage. If you have a 1M DMM, you'd like to see the voltage less than perhaps 0.1% of the voltage on the other side.
What masquerades as a voltage measurement is actually a current measurement done by measuring voltage across the meter's input resistance. For DMMs, that is a somewhat chancier business.
DMMs are handy, but just like an analog meter, you have to know what's inside them, because when you attach them to the circuit, the meter becomes part of the circuit. First off, analog meters were pretty up-front about it - they rated their impedance in so-many ohms per volt. 20K ohms per volt was a normal reading. What that meant was that if you set a 20K 0hms-per-volt meter to 100V, the input resistance was 100x20K, or 2M. This worked out because the analog meter movement was fundamentally a current measuring movement, and everything was scaled from that current-full-scale sensitivity and the movement's resistance. So when you moved the scale to a higher voltage, it necessarily put more resistance in series.
DMMs **may** do this, they may not. They may use more "innovative" means of scaling. The DMM is fundamentally an A-D converter measuring the voltage across some external resistance. They may scale the input with a resistor divider, they may scale it with other means that do not result in an ohms-per-volt, and many I've seen simply quote an input resistance, period. That's usually 1M or 10M, and may or may not, probably not, change with scale setting. You'd have to dig into the schematic of the meter to know, and simple no-name meters may be almost anything.
Testing a coupling cap for leakage, you're looking for the cap to be an open circuit compared to perhaps several megs (i.e., the old 2M-10M of the analog meter setup). With a DMM of unknown input resistance, the measurement gets tricky. It would be nice if you know that the meter is 10M input resistance on all scales. Then you could simply hook the meter between the coupling cap "ground" side and ground, and look at voltage. The ratio of the voltage on the other side of the cap to the voltage you read is the number you multiply the meter resistance by to estimate the capacitor's leakage resistance.
You'd like the voltage you read to be less than 1% or so of the voltage on the high side of the cap, so that if you had a 10M DMM, the resistance would be no smaller than 100 times 10M, or a giga-ohm, which is sufficiently open circuit to make the AIR around the cap part of the leakage. If you have a 1M DMM, you'd like to see the voltage less than perhaps 0.1% of the voltage on the other side.
What masquerades as a voltage measurement is actually a current measurement done by measuring voltage across the meter's input resistance. For DMMs, that is a somewhat chancier business.
Comment