# Why is the order of these serial components changing the current flowing? (Keithley 236)

I have an electronics setup I'm using for research purposes. I'm having some strange behavior so I've tried to boil it down to the simplest working example. Most basically, a voltage is being applied to a device under test (DUT), which in this case is a 10kOhm resistor.

More specifically, I have two Source Measure Units (SMUs), both Keithley 236s. They're both in voltage source/current measure mode, but one of them (SMU1) actually sources a bias and the other (SMU2) just measures the current going through the circuit serially. I do this (rather than just recording the current SMU1 is supplying) because for my real-use case, the DC bias from SMU1 is going to be mixed with another signal by an op-amp, so the current SMU1 is supplying would not actually be the current going through the sample. SMU2 is set up as described on page 1-4 of the manual linked above, in "Measure (current) only" mode. It should basically function exactly as a simple ammeter you can stick serially into the circuit somewhere.

This is fine, but I'm finding a strange problem. If I test two configurations of this circuit at SMU1=0V:

C1:

SMU1---DUT---SMU2---GND

and C2:

SMU1---SMU2---DUT---GND

I get slightly different behavior from the current read by SMU2. C1 gives a fluctuation of +-0.05 microamps, while C2 gives a steady current of ~0.03 uA, with smaller fluctuations around it, but definitely a steady 0.03 uA baseline current.

I'm not exactly worried that I'm not getting perfectly 0 current, because the manual says that in the voltage source range (11V) for the 236 I'm using, it has a 1mV resolution, so if it's a full 1mV across the 10kOhm resistor, that gives us fluctuations/an offset as large as 0.1 uA.

However, what I'm unsure of is, why are the two configurations giving different currents? (one a reasonable fluctuation, the other a strange systemic offset

• wiring layout, grounding, shielding and interference can all affect your results. use of an RF Cap in the right place may help reduce the noise on C1. – Sunnyskyguy EE75 Oct 11 '17 at 22:48
• Hi @TonyStewart.EEsince'75, thanks. The wiring is the same for both configs except for the slight switch, everything else is in the same exact position. All cables are BNCs with the jacket grounded. – YungHummmma Oct 11 '17 at 22:51
• This is very high impedance and prone to EMI , so be careful for loop antenna effects picking up AC E fields. – Sunnyskyguy EE75 Oct 11 '17 at 23:03
• What happens if you swap SMU1 and SMU2? – Bruce Abbott Oct 12 '17 at 4:54