Elk Output expander lost comm

shenandoah75

Active Member
I had had my first output expander (M1XOV) wired up for many months without issues. It's been running sprinklers and other stuff without fail forever (outputs 17-32). I recently (within last 2 weeks) installed an M1XOVR as my second expander in my office/outbuilding. It's for outputs 33-48.

Today, i was in my office around dusk when i heard the relay trip for the christmas lights (ok, shop lights on). However, when i went up to house a short while later, i noticed that none of my christmas lights were on house side. I checked the breakers and the 4 GFI loads as well as the one powering the elk 24VAC wart in a cabinet were up and running.

So, i went back to the elk keypad where it said something to the effect of "output expander lost comm". I trying to reenroll bus modules twice, but both times it failed to find the first one (10 devices enrolled, expecting 11)

The led on the card was blinking twice in short blips followed by long pause before redoing the same...

I shut down the M1 and flipped it back on. This time, the expander was found and i was able to get all my lights up and running (now getting 3 blipss followed by pause). I did no wiring / programming and no control (keypad or webserver) during the last 24 hours since the Christmas lights last cycled without any issues...


any ideas?
-brad
 
I left out one piece of info..

it was the M1XOV (the one with the lower address) that lost comm... It's in line with the second 485 bus (i.e. it is not the terminated card) with the cat6 running into and out of it's connector. So the whole bus didn't fail as all downstream devices (including an input expander and the M1XOVR - the latter is the only temrinated device on that run) were still functioning normally down the road...

-brad
 
Usually these kinds of things turn out to be wiring or termination issues. You can start by checking the bus. With controller off, measure resistance between Data A and B - should be around 70 ohms. Possible intermittent/loose connection?
 
I havent'e measured recently, but have done so on both busses in the past and it was around 68ohms...

1) I'll double check, but the fact downstream cards were fine, i don't think there's an issue
2) the fact that the downstream cards were still working in terms of connection would mean that the incoming/outgoing twisted pairs would have to continue touching each other but somehow not make contact with the card itself. I think that's really unlikely given the nature of the green screw down connectors and the fact the on-board orange led continued to blink?

Can anyone tell me the significance of 2 versus 3 blips of the led on an output expander card? For what its worth, i notice that my M1XOVR has only 1 periodic blip...

-brad
 
Bump..

Brad, have you come to a resolution with this?

I too am seeing this now. I recently converted from a single bus line in one building to two bus lines, each with a DBH. The first bus has an XEP, KPB and XRF. The second bus has an XOVR (as address 2), 2xXIN and 3xPAS. None of the devices are terminated and each DBH has its terminating plug in the first unused jack. I'm measuring about 66 Ohms across A&B.

Occasionally I'll see Expansion Module Trouble/Restore Output Expander 2 messages in the log.

Thanks

Giuseppe
 
Nope - for me it appeared to be a one / two time occurance...

I didn't change the topology / etc at all and christmas lights worked ok the remainder... For what it's worth i use two home runs, not the data bus hubs though... I have one run that is much longer than the other going to my outbuilding underground and over cat6, but the connections are ssolid. And i have both an output / input expander in the outbuilding.

In my case the lost comm was house side in the middle of the run which is why it doesn't make sense to me...

sorry!
-brad
 
Can anyone tell me the significance of 2 versus 3 blips of the led on an output expander card? For what its worth, i notice that my M1XOVR has only 1 periodic blip...

-brad

1 blink is normal operation.

2 blinks mean the expander has no communication with the M1.

3 blinks mean the expander is not enrolled into the M1 system.

4 or more - RUN!!! :(
 
Thanks Spanky!

Interesting - so it's clear the device lost comm as reported so that's cool. I assume Elk's position is that it should be impossible for that conndition to occur if all four conductors were screwed securely to the green connector ? In my case, the 4 strands for each of the four signals (i double up the twisted pair) would have had to have stayed attached to each other but seperated from the green thingy (as none f the comm on either side of that connector was lost)... I am possitive that was not the case.
 
Several things could cause lost communication:

1. Wiring change.
2. Transient damaged the RS-485 driver on the board or some other component.
3. M1 processor reset. This would show up in the log.
4. Data bus noise induced from some device near the wire runs.


Make sure the termination resistors are located at the far ends of the data bus runs, not in the middle of the data bus runs. You should have only two termination resistors installed. The data A and data B resistance should be about 65 ohms +- 10 ohms with the M1 power turned off.

Try moving the output expander back to the M1 and try enrolling. This would tell you if there is some component problem.
 
In my case, the 4 strands for each of the four signals (i double up the twisted pair) would have had to have stayed attached to each other but seperated from the green thingy (as none f the comm on either side of that connector was lost)...
Do you mean you have two conductors going to each terminal. I'm not sure this is a good idea. The twisted pair should be a signal + and signal - as it will reduce noise in the cable run itself. Again, not sure if I grasped your scenario.
 
Ya - actually i grabbed that it from one of the install pics in the showcase. I originally only used 4/8 conductors on each run - pulling back the whites. But i saw this picture (not sure if it was brandon's or someone elses) where people were just pair orange with orange/white in the same wire and i said well it looks better than what i did... so for the board inbetween the M1 and the terminated device, i have 4 conductors in each green connector (orange + white from the prior device and orange + white to the next device) If that's wrong i'll change it back to orange + orange only!

Note i thought this would be an especially good idea for my 100' or so run to the outbuilding so we didn't deal with voltage drop/etc on the +12VDC line/etc...

Ohm wise, i'm good though as noted in past thread... The only other thing i struggle with on a regular basis is the lost data on the serial expanders. I've never solved this, but i don't notice any outward problems except when i lok at the diagnostic menu...

http://www.cocoontech.com/index.php?showto...serial+expander

-brad
 
Back
Top