Heres one for you

[Greg]

I was running the show last night to do some final checking and noticed at 1 unit not working, I figured it was a fuse or something. Today I checked it out and notice now a second unit wasn't working. After checking all fuses, resetting units (since one of these units were in a 32 channel box had to remove one unit to get the other) still no success. Now I am thinking I blew comm chips on both boards which I found very odd.

Network cable well can't be that since 2 units after this cable are still working fine. Here is the cabling run unit 9 (not working) ----> unit 10 (working) ---> unit 11 (workin) ----> unit 12 (not working). Not sure what to do next so I decide to try a new network cable running from unit 8 to the trouble area unit 9, At this point don't think it is going to work. viola... works.

Here is my question, how can the signal not be working on one unit and working on the unit immediatly after?

[Richard Hamilton]

Hi greg,
My question is did you do anything to the controllers while you were changing the network cable? Like, temporarily turn them off?

I can't imagine a situation where a cable can be teh cause if some controllers further down the line are working. Sounds like a controller reset issue or something like that.

[colonel]

Greg wrote:

well can't be that since 2 units after this cable are still working fine. Here is the cabling run unit 9 (not working) ----> unit 10 (working) ---> unit 11 (workin) ----> unit 12 (not working).  Not sure what to do next so I decide to try a new network cable running from unit 8 to the trouble area unit 9, At this point don't think it is going to work.  viola... works.

I am seeing this now in that a unit in the middle is "missing". I put a tester
http://www.amazon.com/gp/product/B000KMHL8M/ref=oh_o03_s00_i00_details
on it and had 1 or 2 of the 8 wires bad. I'm using 500' runs and the cable gets brittle after a season. Gets trampled and things happen. Last year I had a unit that Hardware could not see but it worked fine. I buy 1000' rolls and just make up new cables.

[Greg]

Richard Hamilton wrote:

Hi greg,
My question is did you do anything to the controllers while you were changing the network cable? Like, temporarily turn them off?

I can't imagine a situation where a cable can be teh cause if some controllers further down the line are working. Sounds like a controller reset issue or something like that.

Richard here is a little more info.

Earlier in the day the entire network was fine. In fact I ran a few tests the last couple of nights with no problems. I did nothing since I ran the network test earlier in the day, that is why I thought it was a comm chip issue. It seemed to mirrior a problem i had a few years ago.

I did leave the power on while changing the network cable since I was inside my mega tree (that is where these units are located) and it is a real pain to get in and out.

The led went fom blink to solid as soon as the cable was replaced on the non functioning units.

The big question I had was how could two units found after the bad cable be functioning fine.

[Max-Paul]

I have seen this in a industrial situation. But in our case we had to replace the ADAM-4018 to clear up the problem.

Greg I wonder, do you have a 120 ohm resistor at the beginning and end of the RS-485 run? More important, are you keeping your AC wires (both line and load) away from your Comm leads?

[-klb-]

It is possible that the bad connections allowed some charge to build up on that mid stream controller, such that it was not receiving the data, but that something else was allowing the charge to drain off the later controllers, so they stayed within the signalling voltage allowed. Or it may have to do with where in total cable length the controllers were with respect to the reflection nodes. Or maybe the last two controllers just have a slightly different variant of the RS-485 chip that is more sensitive, and was able to read the signal, even though it was too degraded for the ones having issues.

[Greg]

Network is terminated at end, the controllers are at the end of the network, total network length is around 200 feet with controllers locate at defferent lengths within that 200 feet. all network cables cross any ac at a 90 degree angle, there is one area that the cables do run parallel to the AC for about 30feet, although in the last 4 years this has never created a problem. Well seems to be working now. Thats Good!!!

[Max-Paul]

Agreed, Its good that it is now working. Just made one new cable and I have a high rate of making cables that work the first time. Well this is that one that I am going to have to remake. Bad cables happen.

[Greg Young]

re: I did leave the power on while changing the network cable since I was inside my mega tree (that is where these units are located) and it is a real pain to get in and out.

The led went fom blink to solid as soon as the cable was replaced on the non functioning units



Since it was powered up at the time you changed the cable, I suspect the communication chip may have received a slight surge when you did this. It doesn't take much, which is why I always make sure that no power is running when I change communication cables.

I am more fortunate, as my tree's 4 controllers are at its base, so I can be sure no juice is flowing when I am working on it.

Greg