accounting for burden voltage in my multimeter

[jwm]

I'd like to do some low voltage small current measurements, Although I'd eventually like to spring for a uCurrent, I was wondering whether I could do something simpler and cheaper and junk-partable in the interim, namely put an op amp in the loop in a linear regulator configuration taking its negative feedback from after the multimeter to keep the actual voltage presented to the device at the reference level. Something like the attached schematic.

I am not driving crazy loads, I am testing avr uCs and msp430's current usage under different conditions/voltages and want to verify some of the crazy awesome current consumption claims made by the msp430fr family. The meter I am using is a 5.5 digit HP 3478A which is known to have a particularly bad burden voltage.

Again, I don't need uCurrent accuracy (yet), but I'd like something to at least improve things down at the 1.8-3.6V range. circuit improvement ideas are welcome. Something I was thinking of was ditching the 2n2222 and just using the op amp output directly, I don't have any immediate plans to test over the 40mA or so range (cpus driving LCDs and maybe an LED) so can probably find an op amp that can source that in my junk bin.

[IanB]

If the device current is really low (uA?) then I can't see burden voltage being a problem. Passing a few uA through a 100 ohm meter impedance is only going to drop a few mV. The DUT would hardly notice it.

But assuming the problem is worse than that, do you have an adjustable bench power supply? All you have to do is to put a voltmeter across the load, then wind up the power supply voltage until the load sees the appropriate voltage at its input terminals.

[jwm]

If the device current is really low (uA?) then I can't see burden voltage being a problem. Passing a few uA through a 100 ohm meter impedance is only going to drop a few mV. The DUT would hardly notice it.

But assuming the problem is worse than that, do you have an adjustable bench power supply? All you have to do is to put a voltmeter across the load, then wind up the power supply voltage until the load sees the appropriate voltage at its input terminals.

Yeah, the power supply trick is what I was going to try but I figured it would be nice to have something that "just works" so I can dial up and down the voltage on my reference or fiddle with uploading new firmware without the extra step.

Ah true, yeah when running solely at those low currents it isn't much of a burden. I guess until the current reaches the mid tens of mA range then it isn't going to affect things much. I do plan on doing things like testing PWMing LEDs, which may get up to that range. Conceptually would my circuit help though were I measuring higher current low voltage things?

[dannyf]

I don't have any immediate plans to test over the 40mA or so range

What "burden voltage" at that kind of current levels?

whether I could do something simpler and cheaper and junk-partable in the interim

Yes: stick your meter in there and be done with it.