Honeywell TouchCenter communication issues and lightning control

[Miksu]

Hi,
I'm having frequent (couple times a day) keyboard lost-problem with TouchCenter Plus connected to GD-48 (firmware 7.something). Keyboard diagnostics give connection quality varying between 0% and 100%. Cabling is Cat5 unshielded/ Cat6 shielded, last part to mk8 keyboard is 2-pair shielded Belden. There's short (about 20cm) internal run of untwisted 4-wire between panel and Ethernet module inside panel enclosure, this wire came with ethernet-module. Both ends have 680Ohm resistors and power supply is from panel. Total bus length is between 20m and 30m, closer to 20m I think. There's one Krone-row connector (located in main distribution board), where pairs are connected, pairs are correct, twisting has been preserved and topology is definitely bus, not star or something else. Picture of all devices and bus topology attached.

The weird thing, mk8 keyboard has communication quality of 100% and it's last device on bus after TouchCenter Plus. I have already tried with two different TC+, always same problem. Clear wire is connected on Cat6 between all the runs, haven't connected the shielding to anything, can it cause problem like this? If this is power-problem, I can supply ground level devices from separate power supply with battery backup used for smoke detectors, it has enoguh spare capacity. If problem is caused by shielding, should it be connected at both ends (panel, distribution board), or only single end (which one)?

And final question, I want to control lightning with TC+. I have partially correct configuration, since TC+ now sees outputs in lightning section and they can be toggled. I have set output type to Spare / Lighning / Toggle / Reflex, because there's no fundamental need to control them based on zones. However, outputs are now powered all the time, they don't follow TC+ setting. Is it necessary to set them to follow some zone, like log or something else, that doesn't cause/require real alarm? Or create link? Lights to be controlled are in such locations, that they are used only, when panel is unset or part set.

 

[GalaxyGuy]

With the power off, what is the resistance of a and b with respect to 0v ?

I see this sort of thing a lot when people buy panels from eBay (not from my sales though). The fault is the reason they were swapped out in the first place.

The output control should just work. There's no zone requirement. Try setting 1002 as control and you should hear the relay click when switching.

 

[Miksu]

Got this sorted out, I guess, or at least it's been working for about week without issues. Resistance was around nominal 680 Ohm. Thing that seemed to help was using separate power supply for TC+ and Mk8 keypad. They are supplied now by same PSU as my fire alarms. TC+ gives still low communication quality occasionally in diagnostics, but it's not dropping comletely anymore. Mk8 has still high communication (100%). One of my TC+ was bought from eBay, the later one from security retailer. Got outputs working too, issue was with reversed wiring. Now I can turn on my yard lights and block heater from TC+

 

[GalaxyGuy]

Okay, good. It sounds like there was some voltage drop over your cable lengths. This would be more apparent with the TC, as they consume more current than the regular keypads.

Since you're into controlling outputs, take a look at the new SelfMon vmod - it integrates with things like home assistant and has instant web based keypads and much more.

 

[Miksu]

Okay, good. It sounds like there was some voltage drop over your cable lengths. This would be more apparent with the TC, as they consume more current than the regular keypads.

Since you're into controlling outputs, take a look at the new SelfMon vmod - it integrates with things like home assistant and has instant web based keypads and much more.

This looks interesting, especially possibility to use MQTT. I have surveillance camera system (BlueIris in Windows PC and 4 IP FHD-cameras) that supports MQTT. MQTT-control would make it possible to turn on lights when motion is detected by cameras, start recording on all cameras on burglar alarms etc... And of course, all the convenience things

Edit:
Ordered the SelfMon vmod and Asus Tinkerboard ("classic") for HomeAssistant MQTT-server! My summer residence has multiple buildings and there are two locations where controlling GD is useful, but they are located about 50m apart and there's no spare cable for RS-bus, but there is Cat6 and gigabit ethernet. Of course GX Remote is usable, but it's bit complicated with all user codes etc... And to make it secure and yet easy to use, I can limit access via WLAN (Cisco AP+radius etc).

 

[Miksu]

Did recently some tidying on my Dimension connectors, attached are "before" and "after" pictures, I guess it's easy to see which one is before and which one after. Replacing Krone-row connector and other shady components improved reliability even more, comms quality on TC+ is constantly around 40%, which makes it finally reliable. New connections are made using Phoenix Contact Mini block-row connectors on a 17mm component rail. Mini blocks are very compact compared to full size DIN-rail components and they are very flexible: resistors are installed on component holders, that can be simply plugged and removed as necessary. Non-resistor connector has simple bridge piece, that is also simple to attach and remove. I used Wago 211-connectors for power distribution and to make it tidy, installed them in holder. Mini blocks were by no means cheap, but reliability is more important.

 

[GalaxyGuy]

>> reliability is more important.

There's definitely something wrong with your setup. 40% on the TC+ is really bad. It means that for every 100 packets of data being sent to the touchcenter, 60 are either missing or corrupted. This will make the keypad unreliable and needs to be fixed or you'll end up with full blown tampers for keypad loss when the system is set. The 60% retries will also slow down the bus, slowing apps via Ethernet and other keypads.

From above, it looks like you measured the resistance between A and B and that doesn't inform much. A bus communication at 40% suggests only one side of the differential pair is actually working. I would expect one of the following to be true.

- The A or B line is not properly connected to the touchcenter and open circuit.
- The A or B line is being dragged down to 0v. Confirm this by testing resistance between 0v and A, then 0v and B with the power off. Do this at both the panel and at the touchcenter keypad with the power off. These resistances will give an indication of where the fault is.
- The touchcenter may be faulty.

Here you can see a TC+ and a Keyprox. Both are consistently at 100%.

[GALLERY=media, 106465]IMG_20210430_161821 by GalaxyGuy posted 30 Apr 2021 at 4:24 PM[/GALLERY]