Weird intermittent failure of DSC zone expanders

[TinkerBear]

20 hours ago, MrHappy said:

Rated at 1.7ah 

 

-110mA for the main pcb

-140mA  4 PC5108

 

<whinge> Ok, so you're going to make me do my sums again.</whinge>

 

Using the numbers from the PC1864 install manual...
Main PCB @ 110mA (but this isn't on the Aux output, it provides it)
2x RFK5500 @ 135mA
4x PC5108 @ 30mA
3x Swan Quad PIRs @ 10mA (so far)
1x Eyezon Envisalink 4 @ 65mA

 

Totals to: 485mA on the Aux output, which is limited to 500mA.  Closer than I expected, but within spec.

 

Except... is the keybus power considered part of Aux or not?  The install manual isn't clear.  It doesn't mention a separate current limit for the keybus, so probably.

 

20 hours ago, MrHappy said:

1450mA left for the radio expander(s) detection & sounders.... ect ?

 

what voltage do you have at the expender which plays up

 

have you swapped the expander with one which does not play up & addressed it & fitted in the place of the one which does play up ? 

 

Currently, the zones that fail are assigned to expanders inside the cabinet, so they're less than 30cm of wire from the PCB.  I'm measuring 13.62v at the panel, 13.60v at the expander boards.

 

I've swapped boards and addresses a number of times now.  (Wasn't keeping good track of troubleshooting steps at first.)  The problem has always manifested with zones 9 to 21, regardless of what zone expanders are assigned to handle them.

 

Now I'm thinking if the problem continues to recur, I should try with no expansion boards assigned to those zones.  That hadn't occurred to me before now.

[TinkerBear]

19 hours ago, al-yeti said:

So as mrH have physically swapped board locations rather than just addressing 

 

Yep, I swapped other boards into the positions and addresses of zone expanders 1 and 2 at first, thinking I'd just gotten some broken bits off eBay and it would be a simple fix.  But the failure persisted on zones 9-21.  Then I swapped addresses without moving the boards around - to try to rule out interference from the power transformer inside the case - and still zones 9-21 failed.

 

I'm currently waiting for a recurrence (or lack of one) after resetting the wireless receiver buried in the RFK5500 keypad.

[al-yeti]

So remove one device at a time and leave it off to see if it reoccurs

[TinkerBear]

18 hours ago, al-yeti said:

So remove one device at a time and leave it off to see if it reoccurs

 

Yea, that's a good idea - which absolutely didn't occur to me.

 

I'm still waiting to see if it recurs after I reset the wireless receiver a couple days back - I'm at 43.5 hours since a complete power cycle and nothing has happened, which... doesn't necessarily mean anything yet.

[sixwheeledbeast]

Any fault like this, it's going to come down to isolating things until you pin it down. Starting at a broader scope and drilling down.

As above it's not a usual panel over here so even if it was something common to that panel it's less likely for people in the trade to know about it.

 

I'd want to be using 8 core really and tripling up any supply voltages. Also monitoring the voltages of the PSU over a longer duration.

Is this proper copper cable or CCA?

If I read the OP correctly are the keypads remote from the system and the expanders all in the panel?

Could there be a wire damage or earth fault on the longer wires. Could you stick everything on a bench with 1' of wire between.

Any fault that disappears with just a power cycle never leaves me with confidence or satisfaction.

[al-yeti]

10 hours ago, TinkerBear said:

 

Yea, that's a good idea - which absolutely didn't occur to me.

 

I'm still waiting to see if it recurs after I reset the wireless receiver a couple days back - I'm at 43.5 hours since a complete power cycle and nothing has happened, which... doesn't necessarily mean anything yet.

AC filter needed maybe?

[al-yeti]

11 hours ago, TinkerBear said:

 

Yea, that's a good idea - which absolutely didn't occur to me.

 

I'm still waiting to see if it recurs after I reset the wireless receiver a couple days back - I'm at 43.5 hours since a complete power cycle and nothing has happened, which... doesn't necessarily mean anything yet.

https://www.securitywarehouse.co.uk/catalog/intruder-alarm-equipment-c-36/intruder-alarm-ancillaries-c-36_46/act-230v-mains-filter-p-446.html

[TinkerBear]

It's now been 113 hours without a peep since I reset the wireless module and power-cycled the panel.  (At risk of jinxing it:) I think it's fixed!

 

Thanks to everyone for taking the time to make helpful suggestions!  I've been banging my head on my desk for a month now.  Just explaining the problem to someone who would know what I was talking about was amazingly helpful.  I was thinking this was akin to rubber duck debugging, but as none of you are made of rubber, this is more accurately known as "confessional programming".

 

I'm so happy I don't have to throw more money at this thing.