Resetting relay count

[Les Bartel]

I've been using a PXI-2527 relay card/mux.  As I am still in the development stage of my application, I've not used the card a great deal yet.  However, at one point, I noted that some relays had achieved an activation count of over 90,000.  Now the report lists those relays with an activation count of less than 20,000.  I have not knowingly reset the relay activation count.  I don't even know how to do this.  My concern is that the count has been reset somehow.  Eventually, the relays will wear out, but I won't know when they are reaching the end of life based on the specified rating on closure count and current during closure.

What could have caused the count to be reset and how do I prevent this?

Has anyone developed a testing procedure to evaluate the health of the relays?

[chericks1]

Hi Les Bartel,

What could have caused the count to be reset and how do I prevent this?

This is the first time I've seen this issue.  However, I suppose it is theoretically possible to lose the count quantity you provided.  Please keep in mind this is a hypothetical.  This scenario requires:

a) Only one relay is continuously actuated for a period of time.
b) The computer is shut down or experienced a loss of power immediately after this time.

The relay count on the PXI-2527 relay card/mux module is transferred from volatile memory to non-volatile memory every 10 minutes.  We do not perform this backup more often in order to conserve non-volatile memory lifetime.  The PXI-2527 can perform 140 cycles/second.  If you were only using one relay, 84k actuations would take 10 minutes of constant activity.  This is an interesting number, as you stated you lost around 70k, and this is within the 10 minute time frame.  If you had continuously actuated just one relay, and the computer was powered down before the relay count was backed up, you would lose all the counts that had occurred since the previous backup. 

To prevent this possibility from occurring, I recommend waiting 10 minutes after tests are performed before shutting the computer down.

Has anyone developed a testing procedure to evaluate the health of the relays?

Path resistance may be the best measurement to perform, as it is a clear indicator of when a relay is not healthy and needs to be replaced.  Unfortunately, path resistance can not easily be utilized to test the health of relays during their lifetime.  Please refer to the following document:

Switching Considerations
http://zone.ni.com/devzone/conceptd.nsf/webmain/6f621d198310d75e86256f16001d882d#3

Under Path Resistance, there is a figure showing a typical path resistance of a module with a mechanical life of 50 million cycles.  As expected, the path resistance deteriorates with age and at the end of the relay's lifetime, the path resistance increases exponentially.  However, also notice a discrete path resistance measurement at 20 million cycles may be the exact same as a measurement immediately before the end of the relay's lifetime. 

Therefore, to answer your question, you may want to use path resistance as an indicator of when a relay needs to be replaced.  For relay health during its lifetime, I apologize but there is, as of yet, no easy answer to provide.

Hope this helps!

Chad Erickson

Switch Product Support Engineer

NI - USA

[Les Bartel]

Hi Chad,

The relay closures occurred over a period of several weeks with perhaps a few hundred closures per coding cycle.  So shutdown before the count being written to nonvolatile memory is clearly not the reason for the loss of the count.

As far as testing the contact resistance, I was thinking about a fixture that would present a load (several 10's of milliamps?) and comparing the applied load to the measured load.  But, like you said, a relay can fail at any time.

You didn't come right out and say it, but it does sound like there is no way to reset the closure count in the non-volatile memory.  Is this correct?  Do users replace the relays based on closure count?  Or do they do it upon noting a relay failure?

 - les

[chericks1]

Hi Les Bartel,

You are correct in stating the loss of relay count still does not have an explanation.  I cannot explain what might've happened in this case, as there have been no reported incidents of such a scenario in the past.  You are the first as far as I know.  Perhaps someone else has had this experience and will post as well.

There are several methods to measure path resistance.  The one you propose is exceptable, and could be relatively easy to automate.

As far as resetting the count, you can perform a reset using the Switch Soft Front Panel.  Under File->Relay Maintenance, you can reset and/or edit the relay count stored in non-volatile memory.  Although it depends on the application, most users will replace a relay upon noting a failure.  Relay lifetimes are not exact, as a lot can play a part in how long a relay lasts.  I've known several users in the past who will use one particular relay until it reaches a count limit.  Then, they change to using another relay of that same module.

Let me know if you have further questions!

Chad Erickson
Switch Product Support Engineer
NI - USA