4-wire (4-points) Van Der Pauw measurement (NiDC Power, PXIe 41-45)

[Gordon_s]

Hello everyone, I am extremely new to LabVIEW, so I apologize for my ignorance.

I want to build a program to perform 4-point I-V measurements (ultimately to calculate resistance) using an SMU PXIe-4145. I don't want to perform a single-point measurement, but rather build a ramp by scanning the potential (or current, in my case) on one channel (two physical connections to the sample, V+ and V-) and then measuring, for each point, the current (or potential, in my case) in response (another two physical connections to the sample, I+ and I-) on another channel.

I believe it is possible to use 4 physical connections and just one channel on this instrument—in which case, great! But I still don't understand how to do it. Even using two different channels (one for voltage and one for current) would be fine. I'm using the NI-DCPower drivers, as suggested by the datasheet.

I have checked out the example VIs, particularly the NI-DCPower Hardware-Timed Voltage Sweep. However, from what I understand, this is a very standard two-point measurement setup.

I'm also looking at the Hardware-Timed 2-Channel Sweep (which, from what I understand, performs two separate measurements on two independent channels). I also checked out Ni-DCPower Simultaneous Operation, but I don’t quite understand it.

My ultimate goal is to calculate resistivity and then change the order of the physical contacts to perform measurements using the Van der Pauw method. This means, for example, if I label the sample’s physical contacts as 1, 2, 3, and 4:

• Apply current between 1 and 2 for (e.g.) 60 points and measure the potential between 3 and 4.
• Then apply current between 2 and 3 and measure the potential between 4 and 1.
• And so on.

If anyone can help guide me in the right direction, I would really appreciate it.
Thanks!

[Basjong53]

I can give some insight. A single SMU channel has 4 contacts: Force + and - where the current flows and Sense + and - where you measure the potential difference. This is a 4 wire measurements, and you have to set "Remote" in the sense property of the DCPower api.

You cannot change which contacts are F+, F-, S+ and S-. You will need a multiplexer to physically switch the contacts of the SMU to the contacts of the DUT.

[Gordon_s]

Thanks Basjong for your help.

So, let me see if I got this right: I can perform 4-terminal measurements directly on a single channel (which was my suspicion). To define in LabVIEW whether to use the classic 2-terminal configuration (which I assume is the default if I don't specify anything...?) or the 4-terminal configuration that I need, I have to set the sense property to "Remote" instead of "Local" using the niDCPower Configure Sense VI.

So, "Local" would correspond to a V-I measurement with 2 physical contacts on the sample, while "Remote" would use 4 physical contacts within the same channel.

If that's correct, why is this terminology used?

Thanks a lot for your help!

[Basjong53]

Correct. In "Local" sense (or 2-wire) you would connect the F+ with S+ and F- with S+ like this:

In a true 4-wire measurement, the sense contacts are completely disconnected from the force contacts to eliminate any contact resistance. A general sheet resistance measurement will look like this:

 

A van der Pauw measurement is a special type of 4-point resistivity measurement, but the idea is the same. You do need to change the position of these probes among the corners, which you cannot do by software as I said. You need to physically move them or use a multiplexer with a relay to swap the connection.

Edit: As for the naming, it's just a convention (it's used by Keithley as well, for example)

The technique is commonly used in low-voltage power supplies, where it is called remote sensing, to measure the voltage delivered to the load independent of the voltage drop in the supply wires.

---

source: wikipedia

[tito_89]

To pile on:

You can definitely use a multiplexer card.  Also, you can use some (potentially large number of) relay driver circuits and a USB-6501 to go through your various contact configurations. 

There is an expensive solution here which is to use multiple SMUs, one for each configuration, then only enable inputs and outputs on the SMU you want to use to measure with both "hooked up."  (This may just be two SMUs if you source negative and positive currents on each and ignore the reciprocal connections.) 

For what it's worth, I'm in the middle of building such a system (sadly, not with LabVIEW or NI) with a cheap PLC to do the relay actuation and a single Keithley SMU, plus some extra IO for Hall effect measurements and moving my sample in and out of a magnet.

[Gordon_s]

Thank you so much, Basjong and tito_89!

You have greatly clarified the structure of the problem for me. I'm studying and starting to write some code. It's really interesting, and I plan to test it in 3-4 days when I have direct access to the instrument.

In the meantime, I'll make good use of your advice and update the thread as soon as I start taking measurements, reporting any issues I (inevitably) encounter and sharing the program as well.