Strange problem with SEM zone shunts

[lms]

I have a CM9000 (firmware 7.201) with a LEM01-M2.

Initially there were three SEMs attached (all firmware 7.207) with the first and second SEMs being the older SEM01-M2 boards and the last one being a new SEM01C (but with the old SEM01-M2 power supply board). The first one has a LEM01-M2 but not yet the 2nd or third. Each of the four locations has it\'s own AC power supply and battery all have 0VDC commoned.
The problem with the above configuration was zone troubles on the zones of the 2nd SEM01-M2 - jumpers are set for no EOL resistors. I\'ve removed the cables and have directly connected switches for testing but the problem persists. The zone announcements appear to be working correctly but the zone trouble is also announced. I have not yet configured the last SEM (other than setting the ID to 3). If I bypass these zones, the zones are no longer announced, but the trouble announcements still occur.

I then replaced the first SEM01-M2 with a new SEM01C (but also with an old SEM01-M2 power supply board) and set it\'s ID to 2. The problem persists. Further, on this new SEM01C, the \"Use Shunts\" setting is the default and cannot be changed. The last SEM (also SEM01C) correctly shows \"No EOL\".
All SEMs are being polled but the 2nd & 3rd ones are reporting the Serial Number as \"Invalid\" while the first one reports \"Unassigned\". And the trouble zones appear to be correctly showing Active or Inactive.

This has me rather stumped and I\'d appreciate any advice.

[slychiu]

I am not sure what is happeningrecorded so ignore theseIf the zone reports zone trouble then it should not report on or off at the same time. However zone trouble could be reported in the ON position and restored on the OFF position, so test the zones in On and Off and check if one of the positions is reporting trouble. 

Perhaps the slave power supply is not giving the correct voltage. Check the 12V output is 12 to 13.8V..

Serial number for slaves are not written so ignore these.

the new SEM01C without shunts should have the option No EOL. But you need to do a Scan for modules first for Comfort to detect the SEM type

Please Read the Event log and save it and check which zones cause the trouble. After saving you can erase the event log so you only read new events after that.

I suggest to remove all SEMS except 1 and test for the trouble. perhps 1 of tghem is causing a conflict

[lms]

Thanks slychiu.

Have already tried multiple scans and also voltage checks but hadn\'t checked the event log. Have done so know - nothing obvious - a typical example is...

10/12 11:02 Zone Trouble # RECEPTION_M (36)
10/12 11:02 Alarm Type # ZoneTrouble (5)

So far, I\'ve only had this problem on zones 33 to 40 but haven\'t individually checked each one and haven\'t yet added anything to the 3rd SEM.

The first SEM01C (ID 2) was still only showing \"Use Shunts\" while the 2nd SEM01C (ID3) correctly showed \"No EOL\", but I haven\'t yet connected anything to SEM ID3.

The zone troubles appear to occur when the zone goes Active.

Voltage does appear do be slightly high at about 13.91V with no battery connected. When connecting the battery this goes down to around 13.86/13.87V (don\'t know how accurate the meter is!). Have checked the voltage on two old SEM power supply boards as well as a spare CM9000 - similar with slight variances. Could this high voltage cause a problem?

I have tried the old and a new SEM at the problematic ID2 location but I may also swap the two SEM01Cs (ID 2 & ID 3) and see what happens.

I have now done some further tests...

I left SEM ID2 & ID3 in place but swapped their IDs. After a rescan and write, both ID2 & ID3 SEMs\' zones correctly showed \"No EOL\". I then swapped the IDs back and both still showed \"No EOL\" but the zpne troubles continued. I then rebooted and SEM ID3 zones have gone back to \"Use Shunts\". Am I going crazy?:X:?

Not being an electronics or Comfort genius, maybe I should just order a couple of the new power supply boards for my two SEM01Cs, replace the existing SEM01-M2, upgrade to the CM9001, and hope for the best! Unless someone has some ideas

[slychiu]

Voltage doend s appear do be slightly high at about 13.91V with no battery connected. When connecting the battery this goes down to around 13.86/13.87V (don\'t know how accurate the meter is!). Have checked the voltage on two old SEM power supply boards as well as a spare CM9000 - similar with slight variances. Could this high voltage cause a problem?

That voltage is normal and should not cause a problem 

I left SEM ID2 & ID3 in place but swapped their IDs. After a rescan and write, both ID2 & ID3 SEMs\' zones correctly showed \"No EOL\". I then swapped the IDs back and both still showed \"No EOL\" but the zpne troubles continued. I then rebooted and SEM ID3 zones have gone back to \"Use Shunts\". Am I going crazy?:X:?

In the Command box, can you  type EL and ENTER, let me know what is the reply in the Monitor I/O window.eg> EL
< EL110100FFFF 
This tells us if the Slave and LEM have shunts or not. From the Comfort Protocol document;EL - EOL Shunts Value
The EL reply shows if Comfort, Slave, and each LEM has shunts (1) or is without shunts (0)
>EL
<ELAABBCCDDEEFF

Where AABBCC.. etc are the values of the EOL setting  for Comfort + LEM or  slave and its LEM
AA = EOL Shunts  for Comfort + LEM
BB = EOL Shunts for SEM #1 + LEM
CC = EOL Shunts for SEM #2 + LEM
DD = EOL Shunts for SEM #3 + LEM
EE= EOL Shunts for SEM #4 + LEM
FF = EOL Shunts for SEM #5 + LEM
Each byte of the EOL parameter consist of the EOL shunts status of the Controller + its LEM or Slave and its LEM
Least significant nibble is for Controller or Slave
Most significant Nibble is for LEM
Nibble value = 0 for “No EOL shunts” (Comfort 3), 1 for “with EOL shunts” (Comfort 2)

Eg
EL101100011001 means
Comfort: No shunts (0)
LEM: Has shunts (1)
SEM #1: Has shunts (1)
LEM: Has shunts (1)
SEM #2: No shunts (0)
LEM: No shunts (0)
SEM #3: Has shunts (1)
LEM: No shunts (0)
SEM #4: No shunts (0)
LEM: Has shunts (1)
SEM #5: Has shunts (1)
LEM: No shunts (0)


If SEM is not in the system the value for its shunts byte shall be FF

[lms]

Thanks one again for our assistance.
EL returns EL 11 11 F1 10 FF FF (my added spaces)

CM9000 and SEM01-M2 (ID1) are thus correct (11 & 11)
The first SEM01C (ID2) indicates the SEM is not present and the LEM has shunts - this is a new LEM01B. Doesn\'t seem to make sense!
The 2nd SEM01C (ID3) incorrectly indicates shunts for the SEM and no shunts for the LEM (incorrect as this is an old LEM).


Did another scan and write - no difference.
Reset the system, then got EL 11 11 00 10 FF FF but a short while later this changed back to the above. The \"10\" is also strange - this should be \"01\".

I\'m now sort of wary of connecting any additional boards in case some unknown fault is causing other boards to be damaged!
Am even considering replacing everything, including the two SEM01Cs and LEM01B that I purchased a few months ago - will have to find some budget first! .
But then again, I suppose it could be some cabling problems somewhere. so I\'m planning to take everything down and test on the bench, but would value any further advice you may have.
Thanks for all your help.

[Swiss-Toni]

Possibility the shunts have oxidised and creating bad contact clean them and maybe gently pinch the clips before re-attaching

[lms]

Thanks Swiss Toni:
Being close to the sea, I am in an extremely high rust area so that is definitely something to check.
The newer boards are only a few months old, but the terminals may be the originals going back to the mid-2000s. I may consider replacing the cabling, but these days I tend to struggle with my old back & legs in the roofspace! But I want my Comfort back so will probably have to make the effort!

[slychiu]

It seems to be some intermittent connection problem. I suggest to just connect the CM9000, LEMs amd SEM and their LEMs. 
I also suspect the LEM 10 way flat cables. Perhaps you isolate the module which causes the problem

[lms]

Thanks for your input, slychiu & Swiss Toni. I think I may have found the problem....

I left the CM9000 in place in the roofspace but moved the first SEM (SEM01-M2, ID1) to my work-bench and did all the necessary connections. The two new SEM01Cs (ID2 & ID3) are also still on the work-bench. Still got the same problem so it seems like the cabling in the roofspace is not causing the problems.

I then started swapping around the SEM IDs and eventually the 1st two SEMs (now SEM01Cs on ID1 & ID2) are responding correctly.

The first SEM01C (ID1) has an old LEM01-M2 and is now correctly showing \"No EOL\" for the 1st 8 zones and defaulting to \"Use Shunts\" on the LEM. The second SEM01C (ID2) has a new LEM01B and all 16 zones are correctly showing as \"No EOL\".

The SEM01-M2 which was the first SEM is now the last SEM (ID3). This also has an old LEM and all zones are showing as \"Unknown\" or \"Inactive\". I have tried three different LEMs on this with the same result, so it\'s probably that SEM that is at fault - I guess it\'s causing some problem on the RS485 bus so is affecting all boards down the line. (I haven\'t tried those LEMs on the CM9000 so not 100% sure whether these LEMs are OK but don\'t feel like another crawl around the roofspace!) And all the trouble announcements have disappeared.

I will be ordering some new SEMs and other boards shortly.

Once again, thanks for your responses.

[slychiu]

That seems like a reasonable conclusion as the old SEM may be more then 10 years old. I am glad you managed to work it out