variation in resistance measurement using DMM

[R.V.]

i am measuring very small resistance using 4 wire method. I am using SMU to provide supply to the resistance & one DMM 4071 in series to measure current in the circuit & second DMM across the resistance to measure the voltage drop across the resistance. But the value of resistance is varying from  80 mOhm to 130 mOhm.

I m using SMU as current source (v=50mV & I=0.1mA). 

what should i do to get stable reading? please suggest me the suitable way.

[johnsold]

R.V.,

A 4 wire method is the preferred way to measure low values of resistance as you are probably aware.

Can you post a diagram showing exactly how everything is connected?

50 mV and 0.1 mA is 500 ohms.  What values do you get for 80 or 130 milliohm?

Does anything else in the system change when the variations occur? Are the variations abrupt or continuous? Does it just go one way or back and forth? Even though your power levels seem low, could you have heating occuring in the device being tested?  What is the device?

Lynn

[R.V.]

i am attaching the schematic, so you can understand the arrangement. As i start measurement initially i got values near 80mOhm then suddenly it goes above 100mOhm then slowly it reaches  around 130 mOhm then it vary back & forth between 130 to 140mOhm. I m measuring the resistance of a fuse box wire.

(50mV & .1mA is the open circuit voltage & short circuit current which i set in the SMU )

[altenbach]

You have not attached anything yet...

[R.V.]

schematic

[R.V.]

now i attach a portion of schematic to understand the circuit. I use one SMU 4130, two DMM 4071 & a switch 2531

[johnsold]

Ohm's law: 50 millivolts/80 milliohms = 625 milliamperes.  So you cannot be exciting the specimens with constant voltage.

0.1 milliamperes*80 milliohms = 8 microvolts.  0.1 milliamperes*130 milliohms = 13 microvolts

The thermal EMF of the switch can be up to 50 microvolts >> signal.  Path resistance of the switch is 1 ohm typical.  0.1 milliamperes*1 ohm = 100 microvolts.

If the temperature or the path resistance of the switch changes even slightly, you could see errors of the size you are reporting.

What happens if you connect directly to one sample without going through the switch?

Which current range are you using on the DMM, 100 uA or 1mA?  Is it possible that the DMM is switching ranges?

What are the specimens?

Lynn

[R.V.]

specimens is thin simple wire of length 100mm.

[R.V.]

you are right voltage drop is varying between 8microvolt to 12 microvolt when i use switch. i did the test without using switch & i find very small variation & reading are almost constant. voltage drop vary b/w is 1.9 microvolt to 2.1 microvolt & current vary from 0.105mA to 0.110mA.

Readings are much better without using switch. Is there any way to get these readings using switch.

[Brandon_G.]

Hey RV,

Is there anyway you can apply a larger test current, say 1 mA vs .1 mA?  By doing this you will reduce the impact that the thermal voltage has on your measurement because it becomes a smaller percentage of your total voltage drop in your system.  Also, I would consider using Offset Compensating Ohms before each measurement to null the effect of the thermal voltage. 

Also, is there a reason why you are unable to use the 4-wire resistance measurement on the 4071 directly?

Thanks,

Brandon G