PXI-2530 defies the laws of physics when trying to quantify the relay resistance of specific relay pairs.

[GJM]

I have an issue with a PXI-2530 switch card running in 4X32 Matrix mode. I need to determine the resistance of specific relay pairs within the matrix. I have a PXI-4130 and a PXI-4071 in the same chassis so I am taking 4-wire measurements in a configuration like this...

The resistor symbols represent relays in the switch. Here is a matrix-style representation of the switch routes I am using.(This should look more like the switch soft front panel interface)

So both diagrams represent how I am taking my measurement. I Have externally shorted Columns C8 &C9 (shown in first diagram), I am sourcing 500 microamps of current, and toggling the current source for one positive measurement and one negative measurement, I am measuring the voltage with the DMM . For the measurement above I am measuring .2619 ohms. That should of course be my resistance of relays K8 and K41, as well as terminal block and wiring.

  Here is another diagram. This should be measuring the same two resistors....  

All I have changed is the routing from the source meter. Here is the matrix view...

With this measurement I am measuring .0873 ohms. I am measuring the same resistance as the first example except I get a very different measurement.

  It gets even more interesting. If I were to take 4 pairs of wires, as I am using here, there are 16 possible measurement configurations. I took every one of these measurements and half of them gave me about .25 ohms while the other half gave me about .09 ohms. I tried this on a 2nd chassis and got the same result. My data is in the attached spreadsheet. (My examples above are rows 1 & 5.)  

  By taking the measurement the way we are, all of these measurements should be about the same. I have a constant current source, (I even tried a Keithley in place of the 4130 and got the same result.) and the meter is high impedance.    If I had a schematic for the PXI-2530 I could do a little more analysis to find out why I get different measurements. Is there some clamping diodes on the rows, or columns within the matrix? Something external must be acting upon this circuit. If I can find out what that is, I could possibly determine my relay resistance to a quantifiable level of uncertainty. 

Any help would be greatly appreciated.

Greg

[enderWiggin10]

Hi Greg,

First of all, thank you for taking the time to document the issue as well as you have--this helps us out a lot.  Secondly, you are taking a voltage reading with your DMM, not a resistance measurement, correct?  The DMM will source a test current when taking a resistance measurement which could hinder the accuracy.  One thing to check would be to look at the relay cycles that each of the said relays has encountered.  This can be viewed in test panels on the relay count of each relay.  However, it is still peculiar as to why the circuit would changed depending on the circuit orientation, provided the test current source is held constant. 

Best,

[GJM · Accepted solution]

Thank you Adam for taking the time to review my post. As it turns out my analysis was fundamentally wrong. The spreadsheet/Matrix view was correct but the associated schematic was not. My schematic should have looked like this...

So It seems I was measuring the wire and terminal block resistance via K6 & K39. (I have been using the DMM to measure voltage. the voltage difference divided by the current difference)

I still don't know why I get 2 different readings when changing the routes, but since what I want is the relay resistance I will move on. I can still use the source meter to source current and the DMM to measure the resulting voltage, I just have to do this all on just two rows.

I will list this as resolved and move on.

Thanks again for the reply.

Greg

[Now_With_Underscores]

Here's a diagram of the native topology on the PXI-2530.  Note that there are intermediate relays that allow us to configure the module into various topologies.  It looks like you're measuring the resistance across more relays than you'd otherwise think you were with a conceptual matrix. 

I'd recommend running current higher than 500uA for such a small resistance measurement to reduce the noise in the reading.  Also, note that characterizing a switch module's path resistance once is not sufficient; each time the relay closes, it's possible/likely that the relay will contact in a different area and the resistance will change by some small amount.

It might be better to add a 0.01 Ohm inline resistance to measure the current and also measure across for the voltage if you need an exact runtime resistance.

[gerardo10]

GJM,

I am attempting to do a 4-wire measurement just like you are doing, except I am using a NI PXI-4110 for my power supply and the NI PXI-4070 for my DMM. So far, I have had difficulty writing the software to integrate all 3 devices to work together.

If you have some time, would you be kind enough to maybe upload some example software onto this discussion forum on how you are doing it?

I would greatly appreciate your help!  Thank you in advance.

Gerardo H.

[Frank-L]

Gerardo,

I was wondering what type of application development environment (Labwindows/CVI, LabVIEW, Visual Basic, etc) you are using so we can better assist you in providing an example. I was also wondering if you are also using the PXI-2530?

[gerardo10]

Frank L.,

I am using Labview 8.5. and yes, I am also using the NI PXI-2530 with the TB-2630 terminal block. 

Thank you for your reply

Gerardo H,

[Frank-L]

Just to make sure that we are on the same page, what aspect were you having trouble with? It seems that if you were to have a DCPower task running outputting at your selected values, and your DMM already running you should not have much of a problem measuring the switch relay resistance. As long as the connections are similar to the illustration above and you have closed the connections it should work.