Page 1 of 1

Boiler burns components

Posted: Tue Jul 30, 2019 12:06 pm
by hans.suys
Hi all,

I just purchased the OTGW interface from nodo-shop.nl (smd version) and connected it to my new boiler Bosch GC9000iWM.
The region on the PCB with the diodes is quickly heating up, you can easely burn your fingers. So I disconnected the board to prevent damaging the board.
After inspection Q1 and D9 and D10 are broken and making a short (heating source). I replaced all broken components once but after reconnection the same effect and broken parts.

Is it possible that my boiler can give that much current to blow those components ? Any solutions ?
Maybe the smd versions of the zeners can't handle the high current ?
Any more specs about latest opentherm specifications and max currents ?

Kind regards, Hans

Re: Boiler burns components

Posted: Wed Jul 31, 2019 1:19 pm
by hvxl
The Opentherm protocol calls for a master (thermostat, or in our case the OTGW) to communicate with the slave (boiler) by changing the voltage on the line. The slave responds by changing the current on the line. If you connect something to the boiler connection of the OTGW that is not an Opentherm slave, but for example a boiler that is designed to supply 0.5A @ 24V to power the thermostat, then the OTGW will likely draw too much current while trying to pull the voltage down to the opentherm idle level. It is quite possible that this will blow Q1, D9, and/or D10.

I couldn't find any information on the boiler you mentioned. Are you sure it is supposed to be compatible with Opentherm? If so, maybe it provides different connections for different thermostat types and you connected the OTGW to the wrong terminals.

Re: Boiler burns components

Posted: Wed Jul 31, 2019 2:31 pm
by hans.suys
Hello,
Thanks for the reply.
My thermostat is a Bosch CW400. With a quick search on internet, I can't find if this one is working with opentherm.
But I see the connection is working from OTGW to this thermostat.
In the OT-logger I see the message "Thermostat connected" and some regular messages every second.

The voltage over the terminals when directly connected to the thermostat is 16V (multimeter).
I have an electronic background and measuring equipment at home, but no experience in opentherm protocol. Can I test/check something ?

Kind regards, Hans

Re: Boiler burns components

Posted: Wed Jul 31, 2019 4:55 pm
by hvxl
Based on the symptoms, the assumption is that the boiler delivers too much current. So let's measure that. Since we are not sure what to expect and it's easy to blow up the current measuring circuitry of a multimeter. I suggest to use some resistors. I hope you have some of those lying around.

The maximum current on an Opentherm line is 23mA. Your boiler produces 16V. So if you put a 560 Ohm resistor across the wires from the boiler and measure the voltage, it should at least drop below 12.9V on an Opentherm interface. If it stays at 16V, it is producing 28mA and is definitely not compliant with Opentherm. Actually, an Opentherm line should be idle (low), which means the voltage should drop to 5V or less.

Note that if the voltage stays at 16V, you will dissipate almost half a watt in the resistor. If you don't have a 0.5W resistor, it may be better to use two 0.25W resistors of 270 Ohm in series.

Of course, if you don't have any resistors and your multimeter has a high amp setting (several Amps, rather than milliamps), you can try that (at your own risk).

Re: Boiler burns components

Posted: Mon Aug 05, 2019 9:22 pm
by hans.suys
Hey,
Indeed, looks like too much current... but why?
I did some measurements while the thermostat is connected to the boiler. The default is 15V and every 25ms there are some pulses (length 2.5ms or 7.5ms) dropping to around 10V. And around every second there is a long pulse-train passing the line. If you want I can upload some photo's and video of the screen of my scope.
Most probably looking at the signals over the line, this must be a recent/new opentherm protocol. Also the fact that the OTGW is seeing the thermostat and can read the code from it.

The test with the resistors didn't change much on the line. I started the boiler with a resistor of 560ohm, no thermostat connected. The line still shows the same patterns of series pulses between 15V and 10V. So the boiler is taking initiative to generate data to thermostat.

When looking at the schematic : the OTGW is trying to drop the voltage to 5V (switching Q1), but my lowest signal is still 10V, so probably the current gets too high for Q1 and burns because it can't pull from 10V to 5V.
Can it be possible that my boiler is just working with higher voltages ? A new standard ? Looks like my thermostat can deal with both.
Maybe adapt the interface to work with 10V? So change D11 to 10V, D9 to 9V1, R3 to 680R, R2 to 120R ?
Kind regards, Hans

Re: Boiler burns components

Posted: Wed Aug 07, 2019 1:46 pm
by hvxl
Your tests show that your boiler doesn't communicate using the Opentherm protocol. Changing OTGW components to deal with different voltage levels will not help. The boiler is using a completely different protocol.

Maybe the boiler can be switched to Opentherm via a setting, or by using a different connecting point for the thermostat. Otherwise you're out of luck, I'm afraid.

Re: Boiler burns components

Posted: Tue Aug 13, 2019 9:17 am
by hans.suys
Thanks for your help.

If I receive more information from the manufacturer about the protocol or settings I will update here.
Hopefully I can find a way to control/integrate my boiler in my openHAB-system.

Kind regards, Hans