MicroLogix1400 expansion I/O

We have a Micrologix 1400 that periodically throws a fault code 0071h. The manual defines this as "Expansion I/O Hardware Error" and "The controller cannot communicate with an expansion I/O module."
There are four expansion modules, all snapped onto a DIN rail with end stoppers installed on both sides. The machine doesn't generate significant vibration, but should I expect the fault to happen less frequently if I were to screw the modules to the panel instead of counting on the DIN rail mounting to maintain the integrity of the connection between the modules?

Thanks for that link. Two of the four expansion modules are relay output modules. I'll have to check what sort of loads they're driving.

Can you get access to the AB knowledgebase?

The following is a quote from KB56882;

The 0071h error code is EXPANSION I/O HARDWARE ERROR, the controller cannot communicate with an expansion I/O module. The first two digits should contain the affected slot number but '00' infers that the slot number is unknown and the issue cannot be traced to a specific module.
If the FRN is 3 or earlier, this is a known issue so update the MicroLogix 1400 controller's firmware to FRN 4 or higher.

What firmware do you have?

Out of curiousity, is this the 24V powersupply model, or the 120V? We have two 1400's here that will come up with that fault about once a week. They are 24V powered. I have several 120V powered 1400's here, and have never had this issue. We were leaning towards power drop on the 24v supply. After two years, and no help from the supplier, or allen bradley, we told the supplier to give us two 120V models. We have not installed them yet.

As usual, there are great suggestions here that are worth checking first:
1) Have you clamped your outputs with inductive loads? Diodes work great.
2) Do you have the updated firmware?
3) Is there any chance your 24V power is being compromised by a large 24V load? Have you sized conductors and filtered your supplies?

To add my experience with the ML1400 and that darn 71H error: used dozens and dozens of these on smaller machines as 'remote IO' that would message between each other. Love them. But some machines would come up with this error in spite of ensuring the above three items were handled. Very frustrating.

Then I put a 1400 on a drag boat. Mounted it on DIN rail like all my industrial jobs. Oops: worst case vibration scenario you could imagine. Vibrates like crazy at various frequencies and slams hard on every launch. The 1400 went about 5 seconds in the very first 8 second pass and croaked with this error: 1000HP blown alcohol that can't shift or declutch: nice. My fellow racing buddies thought I was an idiot for controlling the boat with 'all that electronics'.

So, Steve, I did what you asked about. I removed the DIN rail and mounted the PLC directly to a small secondary plate just large enough for the PLC and modules. Then I attached that plate to the main plate with rubber vibration mounts so the PLC is mechanically isolated from the rest of the system. Now that PLC has gone hundreds of passes without the 71H error.

I also started doing the same trick on every machine that I designed: I screw the PLC to a plate and use rubber mounts to mount that plate to the backplane. And I have not seen any more failures of our machines, even though I was convinced the machines saw minimal vibration. Now, admittedly, this is purely heuristic: maybe I over-reacted, its not science as I would prefer, but it is engineering, I think. I could not measure or analyze the vibration on my machines. All I can do is attest to the fact that by vibration mounting the 1400 on its own plate, all our new machines quit coming up with this 71H error that chased me for way too long. YMMV.

I don't think vibration or module mounting would be the culprit here. Those modules connect to the 1400 via ribbon cable and a connector on the front, as opposed to Point I/O or Wago slices that have pressure pins and contacts on the side of the module. At any rate, I highly doubt mounting the modules directly on the panel will solve the problem. I would check the ribbon cables and connectors for damage.

Only ever had this happen once, and that's because the guy assembling and wiring it up did not seat the connector properly. Never an issue since.