PS Sense Terminals and Current Measurement

[colorado.rob]

I am trying to measure the current being drawn from my circuit.

     ------- DMM --------   
   /+                    \
PS                        DUT
   \-                    /
     --------------------

If I run sense wires from the PS to the DUT, the current shown at the DMM drops from 3.5uA to 1.7uA when in STOP mode.  This only occurs in the unit's 200uA range.  In higher ranges, it measures consistently 3.5-4uA. (This is a Siglent SDM3065X, 6.5-digit DMM).

Is that a legitimate thing to run PS sense wires like this, or should the + sense wire go to the DMM.  I thought I understood what sense wires were for, but this result was not what I was expecting.

One issue I ran into was that the PS would into OVP mode when I power on the device (50mA startup, 13-18mA running) with remote sense enabled, but not with it disabled.  That seems to have been an artifact of having auto-ranging enabled on the DMM.

Is this a fundamentally flawed setup?  Can someone school me on this topic?

[BravoV]

Assuming you are attending the whole process and your PSU can be finely adjusted, use above setup with two DMMs and single resistor.

You will get a pretty accurate measurements.

[e61_phil]

I would be carefull with such small currents and sense wires bridging the current meter. Often there are resistors between sense and output for the case you forgot to connect the sense wires. At small currents the burden resistance of your meter might be that high, that the internal connection between out and sense might give significant errors.

Perhaps I haven't completely understand your setup, but the drop in the measured current makes sense if there are internal connections between sense and out. If you switch your meter to higher ranges, the burden resistance is getting lower and therefore most of the current will run through the meter and the internal sense connection gets less significant.

Why do you need sense wires from your PS if you driving just a couple of mA?

[TimFox]

Note that some power supplies with external sensing option install diodes internally between "source" and "sense" terminals to avoid damage when they are not properly connected externally.  This requires that the voltage drop between the terminals at the power supply must be well below the diode turn-on value.

[colorado.rob]

Why do you need sense wires from your PS if you driving just a couple of mA?

I didn't need to.  I was tired and trying to identify & work around a small, unknown voltage drop at the DUT.  The voltage drop was caused by the MCU programmer I had attached -- I was too tired and not paying attention to the obvious.  But, after having identified the problem and resolving it, the sense wires were still attached and I noticed the difference in the current readings on the DMM.  I was not expecting that and came here to be educated.

Is my DUT drawing 1.7uA or 3.5uA?  Are the sense wires helping or hindering in this case?

[Ian.M]

I think I'd add a unit gain OPAMP buffer for the sense line, run from a series pair of CR2032 cells, with their common connected to the PSU side of the current shunt.  Use a low power ultra-low bias current OPAMP.   That should get the current measurement error due to the sense line down to the picoAmp level. 

N.B. if the current measurement circuit goes open or drops more than 3V at any time, its likely to destroy the OPAMP.  You may need a series resistor on its input and low leakage clamping at the input to the PSU side of the shunt to limit the input voltage to +/- 2V to 3V.