Home Automation Domotica Forum Europe, Bwired Forum

Nice work, looking great.

It isn't perfect though, yesterday evening and this evening a gateway crash occured. This resulted in a communications fault, wich forced the boiler in maximum heat demand...

Strange. I've had my first gateway running for weeks without a glitch.
Although that's an older PCB version (November 2011) with the PIC 16F628A and MAX232 connected to Serial > Ethernet server.
And what do you mean with a crash, suddenly no more OT frames being sent by the gateway, or ??

I'm interested, cause I'm also going to ditch the MAX232 and connect the gateway to Ethernet (and do some 'extras' along the way )
Maybe it's smart to see how it performs the way it was meant to be, i.e. with MAX232...?

Digit wrote: And what do you mean with a crash, suddenly no more OT frames being sent by the gateway, or ??

I mean by that no more (valid) frames. If OT communication is halted, the boiler goes to short-circuit state.
PS. yesterday OT communication failed, connected with the PC via a USB BUB. Today OT communication failed connected with the bluetooth module.

So do you get invalid frames then?
What do they look like, can you post some samples?

Digit wrote:So do you get invalid frames then?
What do they look like, can you post some samples?

I guess so, but I haven't got any proof. The gateway didn't log any messages anymore. I should have checked the amount and the type of errors before I pulled the plug. I was in a little hurry because my radiators became red hot and room temperature went up over 22 degrees celsius...

I will give it another try this week.

I am running the OT gateway on the new pcb, including the MAX232, for more than 3 weeks now without any problem. So probably it is not the pcb at fault, but maybe a loose contact or a faulty component. Did you perform the test measurements as described on the OT web site?

Also have the gateway running for 4 weeks now.
server is upstairs in the attic and the gateway down in the metercabinet connected via serial
no problems at all

Templar wrote:It isn't perfect though, yesterday evening and this evening a gateway crash occured. This resulted in a communications fault, wich forced the boiler in maximum heat demand...

I haven't had any reports from other users with issues like this. And even if the code somehow took a wrong turn somewhere, the watchdog timer should kick in and put the device into a fail-safe state where no code is required to allow communication between the thermostat and boiler.

I did get a suggestion from someone that it would be good to put a 0.1 uF ceramic decoupling capacitor between the power pins of the PIC. Without it, a bad or aging C2 could cause erratic behavior. As you appear to be experiencing erratic behavior, you may want to check your C2 (correct value, correct polarity, proper soldering) and possibly add such a decoupling capacitor.

Just moved the gateway to the attic and fiddled around with some google graphing and some perl code:

My gateway would occasionally fall back to fail-safe mode due to a watchdog timeout. Even though nobody else reported any similar problems, it bothered me. The problem was hard to diagnose because it happened so infrequently, but I think I have found the cause and fixed it in firmware version 3.3.

Knowing the cause of the problem now I'm even more surprised nobody else reported running into this issue. I suggest everyone using a PIC16F88 based gateway for more than just monitoring purposes to upgrade.

Got my Opentherm Gateway operational now.
Also created a webpage for it, to show (off) how it's performing, real-time (not with IE).
Finally I have both OpenTherm and the means to control the thermostat from my system

A really big thanks to Schelte for all his work!

Hi Digit,

Do you have the GW now connected via Ethernet or still via RS232?

Ethernet; here's a picture on which you can see how.

Nice job! a bit too much for me, I'll stay with RS232