High Resistance in 176-pin cable for PXI-2530

[Jeffrey_Zola]

I am using a pair of PXI-2530 modules in the 4 wire/32 channel multiplexer topology. My DUTs are connected to TBX-50's that are connected to the modules with the 176-pin cable. I have the original PXI-2530 modules, not the updated version.

I am using the MUX to measure voltage using one of the differential pair signals and current with the other. The other day, I noticed that one DUT out of 60 had lower current measurements than the others. I used a hand-held DMM to identify that there was a higher resistance between the COM- terminal and that DUT's negative current terminal- putting the DMM probes on screw #34 and the screw for the DUT's connections. All of the channels I checked- both positive and negative side- gave about an Ohm between the screws with its MUX channel active, except this one which was about 50 Ohms. I moved the 176-pin cable to an unused module and verified that the same higher resistance is present. I also flip-flopped P2/P1 and the higher resistance moved with the cable branch, thereby eliminating the terminal block as the culprit. So I swapped the 176-pin cable for another. I measured the resistance between the COM- terminal and the negative current terminal with its MUX channel active. It was now about an Ohm, and I was happy, although I had a suspect MUX cable.

Today, I noticed that two other channels (same MUX) were exhibiting similar behavior. I ohmed out the screw terminals- and saw that one of the channels appeared to be in the kOhm range, but did not swap the cable because I decided to let the experiment run for a while to collect more data before making decisions. Plus, it was about 3PM and I haven't had lunch. Or breakfast. And was wondering if I was hallucinating.

Does anyone have experience with high resistance in specific connections of these 176-pin cables? Are these not intended for applications where such changes in resistance can affect measurement accuracy? I saw the 50 Ohm resistance in the first cable by a slightly lower current through a 100kOhm DUT. The current channel that measured in the KOhm range is about 15% of what is expected.

Thanks for the help!

Jeff Zola

[Now_With_Underscores]

Hey Jeffrey,

Sorry you're having problems with your switching setup.  No, the 176 pin cable should not exhibit high resistance.  From your experimentation thus far, it sounds like you have at least one failing cable (the first one that followed the cable in your initial experimentation).  It's unclear whether or not the second high resistance you've described is due to the cable or the module; I recommend measuring the resistance of just the cable with a DMM to verify both events are the same issue.  With a DMM connected to each end, move the cable around and see if the resistance appreciably changes. 

How are these cables mechanically routed in the system?   For example, is there a small bend radius that would strain the outer wires inside the core?  Are the cables radially twisted multiple times?

[Jeffrey_Zola]

Thanks, John. I do not believe it is the module that is causing the higher resistance with the second cable, as it is the same module that was connected to the first cable!

Not sure how to realistically measure the resistance of just the cable, as the connectors on the 176-pin side are small and the D-Sub's are females. I supposed I could locate a set of miniature DMM probes. I have chosen to hook up the cables to different modules as a way to partition.

I probed the 3rd cable again today, but did not see the resistance in the same range as before. I left a test run active with the 3rd cable to see if anything is visible in the data. It may be that there was a loose connection or there may be a misaligned pin on one of the connectors.

The cables are not stretched or kinked significantly. The PXI chassis is on a rolling cart, positioned on a shelf that is a typical work surface height from the floor. The terminal blocks are suspended from the side rails of system I am evaluating, and 16-18 inches above the height of the chassis. The 1m cables make a fairly gentle S-curve as they extend from the front panel connector up to the terminal blocks, which I have attached to each other such that the cables extend in opposite directions away from the block assembly. And each pair of MUX terminal blocks has a terminal block that is connected to a PXIe-2569. There is a massive mess of wire and cable!

Jeff

[Jeffrey_Zola]

I have some more information on this issue after taking time to perform some pretty tedious measurements.

I have three PXI-2530 modules, appropriately called "MUX1", "MUX2", and "MUX3" in my system.

I used three of the 176-pin cables. In the experimental setup that triggered my original post, I used two of the MUX modules (#1 and #2). I labeled the cables that were attached to the MUX modules as "MUX1" and "MUX2" and use those in the data summary below. I label the data for the third cable of my original post as the "suspect" cable.

In the experimental setup that spawned the original post, P2 and P3 were used for Voltage measurements, and P1 and P4 for current measurements. I label the data with those conventions as well.

Procedure:

I took one TB-50 and used it for all of the resistance measurements, which were taken with a Fluke 77 handheld DMM. I iteratively measured the resistance for each channel of all three MUX modules with all three cables. I used a topology (16X1 1W) to ensure that all measurements could be taken on a single terminal block, although my original setup used the 4W 32-channel topology. I jammed one probe of the DMM into the port behind either screw 33 or 34 of the terminal block because I only have two hands. The other probe was touched to each screw of the terminal block while its corresponding channel was selected on the soft front panel and I recorded the resistances from the DMM to the nearest tenth Ohm.

I somehow managed to skip one group of the measurements using the suspect cable on MUX2. Either that, or I neglected to save the data. Anytime I took a measurement that appeared to be above the "baseline" of about an Ohm, I wiggled the cables. In most cases where an single measurement was above about 2 Ohms, the readings varied widely with the motion, sometimes pegging the scale of the meter. In the cases where groups of adjacent channels all measured above this 2 Ohm threshold, however, those were unaffected by cable motion.

The attached graphic is my best approach of presenting the results in the least-busy graph. All three cables show high resistance channels. By my luck, most appear on the P1 and P4 legs, which my conventions have assigned to current measurements.

I also compared these results to the "problematic" channels I noted while preparing the original post. In each case, the problematic channels showed up in these measurements, with the one I noted in the KOhm range on my second post being one that pegged the DMM on these tests.

Summary questions:

1. What is the expected range of resistance measurements of this type of test? I recognize that the switch itself has a resistance of a few tenths of an Ohm, as do the probes and the wires. I am not concerned with variation in the ball park of 1-2 Ohms total, but when I see groups of adjacent channels having higher resistances on each MUX module  (MUX1 Cable, P4 channels 16-31, SUspect Cable, P1 Channels 0-15) I suspect manufacturing variation.
2. How many channels per cable are acceptable to have the higher resistance that varies as the cables are moved?

Thanks for reading!

Jeff

[J-Barticus]

Hi Jeff,

In your previous post you mention data/graphics that I believe you meant to attach, would you mind posting with that information? If I misinterpreted your intention to post the results then you can ignore my request.

---

@Jeffrey_Zola wrote:

1. What is the expected range of resistance measurements of this type of test? I recognize that the switch itself has a resistance of a few tenths of an Ohm, as do the probes and the wires. I am not concerned with variation in the ball park of 1-2 Ohms total, but when I see groups of adjacent channels having higher resistances on each MUX module  (MUX1 Cable, P4 channels 16-31, SUspect Cable, P1 Channels 0-15) I suspect manufacturing variation.

---

From what I have gathered from your post, when you see the groups of adjacent channels having higher resistances you are seeing resistance values of 2 ohms, possibly slightly higher. My interpretation from those results leads me to believe those groups of adjacent channels maybe the channels more actively used than the others, potentially nearing their "end-of-life". In the specifications manual for the PXI-2530, there is the "initial" path resistance of less than 2 ohms and the "end-of-life" path resistance of greater than or equal to 3 ohms. My theory/assumption could be verified by check the relay count of the PXI-2530 to see which channels are being switched the most.

For the channel you had that pinged the DMM after physically moving the cable, I believe there is a contact issue for that specific channel on the cable.  As you mentioned this particular channel has the same behavior when this particular cable is used with any of your PXI-2530 modules, so I think it is safe to say this is a failing cable, for at least that channel.

---

@Jeffrey_Zola wrote:

2.    How many channels per cable are acceptable to have the higher resistance that varies as the cables are moved?

---

I'm not sure I understand this question completely, could you clarify? How much is the resistance varying when the cables are being moved? Is this inclusive of the channels that ping the DMM?

-Jake B.

[Jeffrey_Zola]

Hi, Jake. Here is the attachment. Oops!

The X-axis of this pivotchart is broken down by (1) Cable, (2) channel (1-32 instead of 0-31) and (3) MUX Module.

If I had bothered to include the atatchment yesterday, you would have seen that the blocks of higher resistance go with the cable and not the MUX module. Not to say that there aren't potentially aging switches in any of the modules, but I do not think the data points in that direction.

The resistance of many of the channels varied as much as the 10-20 Ohms that was charted into the KOhm range- plus the several that pegged the meter. My question was related to the NI tolerance for such channels in these cables. Are these cables 100% tested prior to shipment, or did my cables escape detection because of sampling plans that may be used?

Regards,

Jeff

---

@J-Barticus wrote:

Hi Jeff,

In your previous post you mention data/graphics that I believe you meant to attach, would you mind posting with that information? If I misinterpreted your intention to post the results then you can ignore my request.

---

@Jeffrey_Zola wrote:

1. What is the expected range of resistance measurements of this type of test? I recognize that the switch itself has a resistance of a few tenths of an Ohm, as do the probes and the wires. I am not concerned with variation in the ball park of 1-2 Ohms total, but when I see groups of adjacent channels having higher resistances on each MUX module  (MUX1 Cable, P4 channels 16-31, SUspect Cable, P1 Channels 0-15) I suspect manufacturing variation.

---

From what I have gathered from your post, when you see the groups of adjacent channels having higher resistances you are seeing resistance values of 2 ohms, possibly slightly higher. My interpretation from those results leads me to believe those groups of adjacent channels maybe the channels more actively used than the others, potentially nearing their "end-of-life". In the specifications manual for the PXI-2530, there is the "initial" path resistance of less than 2 ohms and the "end-of-life" path resistance of greater than or equal to 3 ohms. My theory/assumption could be verified by check the relay count of the PXI-2530 to see which channels are being switched the most.

For the channel you had that pinged the DMM after physically moving the cable, I believe there is a contact issue for that specific channel on the cable.  As you mentioned this particular channel has the same behavior when this particular cable is used with any of your PXI-2530 modules, so I think it is safe to say this is a failing cable, for at least that channel.

---

@Jeffrey_Zola wrote:

2.    How many channels per cable are acceptable to have the higher resistance that varies as the cables are moved?

---

I'm not sure I understand this question completely, could you clarify? How much is the resistance varying when the cables are being moved? Is this inclusive of the channels that ping the DMM?

-Jake B.

---

[J-Barticus · Accepted solution]

Jeff,

Thanks for posting your test results and clarifying your earlier question. The cabling is purchased from a 3^rd party vendor, so I am not familiar with how thorough the testing is for the cables. In your case, it does appear that there is a defect or bad soldering within the suspect cable.

As for variance in the measured resistance caused by movement of the cable, it is difficult to speculate on an expected difference. The testing/specifications are for the use case that the module is screwed into the PXI slot, the cable is attached/hooked in, and left in a ‘motionless’ state during measurements . What I can say is that continuous movement or a high amount of continuous pressure a single direction is not good for the connection of your system as you could be connecting/disconnecting the contacts between the cable and PXI module.

-Jake B.

[Jeffrey_Zola]

Thanks, Jake. My intention is to use 2630 blocks with header connections where possible instead of using these cables and screw terminal blocks.

I am in the process of procuring cabling to interface the equipment I am evaluating to the PXI modules using interface PCBs that I designed and had produced. It has been an uphill battle to have the cables designed and produced, as there are four custom cable assemblies involved to connect to the 2530, 2569, 4071, and 4141 modules, each with its own connector specifications. But I do think I am in the home stretch!

Jeff