11-15-2013 11:42 AM
Does AC voltage present on a "supposedly" isolated and powered-down circuit system present a problem for a DMM trying to measure the resistance of the circuit? How and why does it affect it?
I have read that DC voltage present on a circuit will definitely alter the DMM measurement (correct me if I am wrong about this as well). I use a fluke 753 and a fluke 87 to do my measurements. I understand that a DMM uses ohms law by sending out a current and reading back a voltage, then calculates the resistance (R=V/I)
If I need to clarify what I am trying to do then see below: (sorry if its a long read, just trying to put the devil in the details)
I am trying to verify that my circuit grounds are indeed isolated from each other by doing a continuity test with power disconnected at my power supply (24V DC) between each of their ground buses. If the resistance is high then I know that the buses are isolated. If the resistance is low then the two things I am measuring has a short between them somewhere. (the ground buses under test are disconnected from their respective ground rod wires that physically go out into the earth). Each ground on my circuit is described as follows: Devices on my circuit are either tied to a Protective earth bus bar (PE, for devices that have a decent current draw), an instrument Earth bus bar (IE, for devices that are suseptible to any noise), or a Frame Ground bus (FG, mainly to protect personnel when touching large metal objects like a frame of a device: ie. motor, control cabinet, battery cabinet).
I run into situations from time to time where my circuit is receiving some voltage from somewhere outside my system power supply (be it through electromagnetic induction of nearby wires with voltage present, etc.) and I try to track it down to the best of my ability. However sometimes I cannot eliminate it and have to continue with my isolation checks. The AC voltage present is typically between 250mv to 6 VAC (most often around 1-2 VAC). Typically I notice that when measuring resistance on such a circuit with my DMM that resistance exibits a "growing" behavior. Starts in the kOhms range and grows to MOhms.
Any explanation would be appreciated.
11-15-2013 05:45 PM
Hello presumably fellow Texan,
There are various methods to measure resistance even with a DC bias applied. See Offset Compensated Ohms (OCO), 6-wire resistance, etc. However, I won't go into detail here, as those articles are pretty self-explanatory.
Of course, if you only take a single measurement, and whilst taking that measurement there's any foreign voltage potential applied, your measurement will be incorrect, regardless of whether an externally applied signal is AC or DC... the only difference is that AC signals might return different values on subsequent measurements, whereas subsequent DC signals would be the same. As you mention, resistance is calculated with V/I=R, so any external current or voltage influence is going to disrupt your measurement. However, if you know that the external signal's impedance wherever you're measuring in-circuit isn't insanely low and is constant, you can use OCO to determine the actual resistance.
As an aside, you can actually use this theory to accurately measure the current through a wire, without knowing anything about that wire's resistance... all you do is measure the voltage across the wire, then apply a relatively large current across the wire and measure the voltage again. Take the delta voltage and divide by the delta current and you'll get the wires resistance. You could then divide the initial voltage reading (or take another one) and divide it by your known resistance to determine the (previously) unknown current through the wire. FYI: in real world applications, there are typically two known currents run through the wire, so that if there was initially no current running through the wire, your DMM measuring the voltage isn't buried in the noise floor... but I digress; back to your questions....
It sounds like you have an isolated supply and are trying to determine if the two isolated sections are truly isolated. Since the grounds are unique, and assuming the power rails aren't also shared, the correct test to perform to determine that a device is truly isolated is called 'HI-POT'... what you do is take your unpowered device, short the ground on the nonisolated side to earth ground, then connect a power supply between earth ground and the isolated ground. Finally, turn the power supply up to a few hundred volts and measure the leakage current (ideally 0A). Unless you like fireworks, it's a good idea to use a power supply with a programmable current limit... in case your design fails HI-POT. If there's any uncertainty in what I've described, please ask for clarification, but in general I recommend searching The Google for "high voltage CAT testing" for more information.
The problem with trying to test HI-POT with a DMM's resistance function is that the drive current is very low... typically less than 1mA. Also, the Vmeasure limit is typically 10V or less. Both of these values are too low for a true HI-POT test.
If you're not performing a HI-POT test, please provide a schematic so I can comment further.
Also, the "growing" resistance you measure means your DMM's output current source (1mA, etc) is slowly charging capacitors on your DUT... as the capacitors charge, they consume less current, which equates to a high resistance measurement (R=V/I).
Certainly let us know if you have other questions.
11-17-2013 08:59 PM
Thanks for the detailed response. That is exactly what I was looking for. I appreciate you sharing your expertise.
(Lived in Texas for a few years, hence the name Lonestar, heh heh).