No thermostat connected, but it is

[maartenp]

Recently the transformer of my OTGW gave up. I've replaced it and everything seems to be working (boiler side), but whatever I'm trying, otmonitor keeps reporting Thermostat not connected. Also the diagnostics firmware does not register anything from the thermostat. I'm sure that the thermostat itself is ok because when connecting it without the gateway inbetween, everything works.

Visual inspection of the circuit board and components does not show anything out of the ordinary. Any idea where to get started in figuring out which part is broken?

Thanks for any help you can offer!

Best regards,
Maarten

[hvxl]

The OTGW normally only reports Thermostat disconnected once. What do you mean it keeps reporting it?

What was the result of diagnostic firmware test #5? And what about test #4 (with the interfaces looped)?

[maartenp]

By keeps reporting it, I mean that I can't get the OTGW to detect the thermostat. So power cycling, resetting, reflashing firmware, nothing works.

The requested outputs from the diagnostics firmware:

Code: Select all

Opentherm gateway diagnostics - Version 1.3

1. LED test
2. Bit timing thermostat
3. Bit timing boiler
4. Delay symmetry
5. Voltage levels
6. Idle times

Enter test number: 5
Reference: 1.42
Thermostat: 4.99
Boiler: 0.00, 2.41
Reference voltage setting (0..9) [3]: 3

Opentherm gateway diagnostics - Version 1.3

1. LED test
2. Bit timing thermostat
3. Bit timing boiler
4. Delay symmetry
5. Voltage levels
6. Idle times

Enter test number: 4
OK1A high-to-low: ### Error
OK1A low-to-high: ### Error
OK1B high-to-low: 5us
OK1B low-to-high: 6us

I also ran test number 2:

Code: Select all

Opentherm gateway diagnostics - Version 1.3

1. LED test
2. Bit timing thermostat
3. Bit timing boiler
4. Delay symmetry
5. Voltage levels
6. Idle times

Enter test number: 2

I cancelled this test after one minute by pressing ENTER because nothing showed up on screen, even when restarting the thermostat or changing the room setpoint. Thanks for having a look!

[hvxl]

It must be R6 then. Test #4 shows that the thermostat interface is able to change the level of the opentherm line. So Q2, Q3, Q4, Q5, R7, R9, R11, and R12 must all be fine. Test #5 shows that the A/D converter is maxed out when measuring the voltage on the thermostat line. So R5 must be pulling pin AN0 to some voltage at or above Vcc (probably even 8 - 24 V if the protection would not kick in). Normally R6 should prevent it from getting any higher than around 3V. But that's clearly not happening. So make sure that R6 is connected to GND on one side and pin 17 of the PIC on the other side.

[maartenp]

I just got around to testing your conclusion. Unfortunately, R6 seems to be doing just fine. Connected to GND on one end (testing no resistance between the same end of R4, also tested connection to D1 and D2 successfully. Also pin17 of the PIC connects to the other end of R6. Testing the resistance between GND and pin17 of the PIC also gives me 4.64K, which seems to be just fine as R6 is supposed to be 4.7K. Any other tests I could be doing? Thanks again for your time!

[maartenp]

It seems that my measuring action has fixed the issue. Could be a loose soldering connection I guess, because after I tested with the multimeter, all of a sudden the diagnostics test (4) starts working. I guess tomorrow I will resolder R6 just to be sure and then reconnect the boiler and thermostat. Thanks! I would never have found this without your help!

Best regards,
Maarten