No boiler communication

This Forum is about the Opentherm gateway (OTGW) from Schelte

Moderator: hvxl

Post Reply
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

No boiler communication

Post by AlfredB »

Hi there,

I've soldered the OTGW together and I'm having some issues on the X2 interface. The IC1 runs and I have loaded the diagnostics firmware. Initially when I connected the thermostat and boiler, the thermostat test with the diagnostics firmware reported timings, the boiler did not. Also running test 4 reports ### Error: Interfaces don't appear to be looped. When I measure on Voltage on X1 is 27V current measures 0,632V with 100 ohm resistor. With pin 3 of IC1 connected to ground (IC1 pin5) current is 1,82V (again with 100 ohm resistor). Now with OK1 placed in it's socket and X1 looped to X2 voltage is 17,73V on X2 17,83 on X1. Now when I connect PIN2 of IC1 to PIN14 nothing changes. Í have tried different optocouplers and also moving the OK1 two pins left or right does not make any difference.

I've measured al resistors and all seem ok.

Anyone an idea where to look?

Thanks!
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: No boiler communication

Post by hvxl »

First of all: You should not have IC1 in its socket when connecting pin 2 to pin 14 of the IC1 socket. (Apparently that's not clear to everyone.)

Your symptoms sound like something is wrong with D11 or R8. Wrong value, bad connection or broken. Check the resistance between OK1 pin 1 and pin 2 of the IC1 socket. If that is not around 1500 Ohm, then R8 is the culprit. With X1 looped to X2, OK1 in place and IC1 removed, connect pin 2 of the IC1 socket to pin 14 and measure the voltage across D11 and the voltage between pin 7 and pin 4 of OK1. If the last two measurements differ, there's a bad connection on D11. If they are the same, but exceed 4.3V, D11 is wrong/broken.
Schelte
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

Re: No boiler communication

Post by AlfredB »

Hi,

First of all, many thanks for responding. I'm aware that the measurements should be done without IC1 in it's socket (read the previous posts about it also :-)).

Meanwhile I detected a broken connection between d11 and OK1, fixed this now but this does not solve the issue. I'll continue checking the things you mention in you reply and post back.

Have a nice Christmas!
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

Re: No boiler communication

Post by AlfredB »

Ok, I checked the values:

Check the resistance between OK1 pin 1 and pin 2 of the IC1 socket. If that is not around 1500 Ohm, then R8 is the culprit. => 1103 Ohm.

With X1 looped to X2, OK1 in place and IC1 removed, connect pin 2 of the IC1 socket to pin 14 and measure the voltage across D11 and the voltage between pin 7 and pin 4 of OK1. If the last two measurements differ, there's a bad connection on D11. If they are the same, but exceed 4.3V, D11 is wrong/broken. => Values are the same, 0,92V

Now I have a question:
So I should measure 1500 Ohm, as listed on the schematic. In my situation R8 is a 1.2K resistor (as listed on the assembly sheet) (I measure around 1100 Ohm). This is basically what's between OK1 pin 1 and IC1 pin 2 also around 1100 Ohm. Should R8 than be a 1.5K resistor?
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: No boiler communication

Post by hvxl »

1103 Ohm is OK, but 0.92V isn't.

Please measure the following voltages relative to OK1 pin 4 (X1 and X2 still looped and IC1 socket pin 2 connected to pin 14):
  • OK1 pin 8
  • Cathode of D10 (side with the ring)
  • Cathode of D9
  • Both sides of R1
For completeness also measure the following relative to IC1 pin 5:
  • OK1 pin 1
  • OK1 pin 2
  • IC1 pin 2
Schelte
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

Re: No boiler communication

Post by AlfredB »

Ah, ok, thanks. What value should it be normally?

Please find the measured values below:

Please measure the following voltages relative to OK1 pin 4 (X1 and X2 still looped and IC1 socket pin 2 connected to pin 14):

OK1 pin 8 => 0.92V
Cathode of D10 (side with the ring) => 19,78V
Cathode of D9 => 1,63V
Both sides of R1 => 19,95V | 20,8V (side of D1)
For completeness also measure the following relative to IC1 pin 5:
OK1 pin 1 => 8,56V
OK1 pin 2 => 0V
IC1 pin 2 => 8,67V
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: No boiler communication

Post by hvxl »

Are you asking what voltage to expect across a 4.3V zener? Surprise: 4.3V.

Now regarding your measurements: If you look at the schematic, the cathode of D10 is supposed to be connected to OK1 pin 8. If you find vastly different voltages at those two points, that apparently isn't the case. So please check the connections highlighted in the images below.
D10 -> OK1 pin 8
D10 -> OK1 pin 8
alfredb1.png (8.38 KiB) Viewed 13289 times
GND1
GND1
alfredb2.png (8.38 KiB) Viewed 13289 times
But actually almost all of your measurements show wrong values. These are the expected voltages:
  • OK1 pin 8 => 4.5V
  • Cathode of D10 => 15V
  • Cathode of D9 => 1.63V is OK
  • Both sides of R1 => 15V | 15.6V
  • OK1 pin 1 => 1.2V
  • OK1 pin 2 => 0V is OK
  • IC1 pin 2 => 5V
It certainly looks like you have more than one problem.
Schelte
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

Re: No boiler communication

Post by AlfredB »

Ah, should have checked the schematic before asking :-)

D10 -> OK1 pin 8
is connected

GND1
all are connected

I'll try to check all connections I can see on the pcb.
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: No boiler communication

Post by hvxl »

Thinking about it a bit more, the biggest problem may be around OK1 pin 1 and 2. The LED in the opto-coupler should be on, in which case I'd expect around 1.2V between the two pins. You reported 8.56V. That means there is either a bad contact, or the LED is fried. In both cases, the output of the opto-coupler will probably not be on. That may explain at least some of the unexpected voltages. Though it doesn't explain how you can get a difference of 19V between two points that you also claim are connected.
Schelte
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

Re: No boiler communication

Post by AlfredB »

Hi, thanks again for your answer.

Meanwhile I replaced the optocoupler with one I "borrowed" from another device.

Now values are totally different:
OK1 pin 8 => 5,98V (should be 4,5V)
Cathode of D10 (side with the ring) => 5,98V (should be 15V)
Cathode of D9 => 1,74V
Both sides of R1 => 5,99V | 6,93V (side of D1) (Should be 15 | 15,6)
For completeness also measure the following relative to IC1 pin 5:
OK1 pin 1 => 1,71V
OK1 pin 2 => 0V
IC1 pin 2 => 6,97 V (Should be 5V)
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: No boiler communication

Post by hvxl »

Sorry, some of my "should be" values were for when IC1 pin 2 is not connected to pin 14. The fact that I put different values for OK1 pin 8 and the cathode of D10 should have been a red flag, as those two are connected together. Here's a corrected "should be" list:
  • OK1 pin 8 => 4.5V
  • Cathode of D10 => 4.5V
  • Cathode of D9 => 1.74V is OK
  • Both sides of R1 => 4.5V | 5.1V
  • OK1 pin 1 => 1.2V
  • OK1 pin 2 => 0V is OK
  • IC1 pin 2 => 5V
So your values now look pretty good, albeit not perfect yet. Based on your latest values, the voltage on X1 will be a bit too high for a proper low opentherm level. It's probably around 8.5V, while it is supposed to be 7V max. But it may work. Is OK1 pin 7 (or the voltage across D11) now 4.3V?

Is the new opto-coupler the specified type? It seems to user higher currents/voltages for its LEDs. That could also mean that the timings may be off. You could try running test #4 of the diagnostic firmware.
Schelte
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

Re: No boiler communication

Post by AlfredB »

Hi, well I'm happy there's some improvement :-) I've bought myself a brand new optocoupler today, the best I could get is an ltv-827 (like this one https://www.conrad.nl/nl/fototransistor ... 87003.html). You know if that one should be ok?

The values (with IC1 pin two connected to pin 14) are now slightly different:
OK1 pin 8 => 5,52V (Should be 4.5V)
Cathode of D10 => 5,55V (Should be 4.5V )
Cathode of D9 => 1,68 (Assume ok)
Both sides of R1 => 5,55 | 6,42 (Should be 4.5V | 5.1V)

OK1 pin 1 => 0,54V (Should be 1,2V)
OK1 pin 2 => 0V is OK
IC1 pin 2 => 5,48V (Should be 5V)

You were quite right about the voltage on X1, it's 8,45V now.

D11 unfortunately (and Ok1 pin 7) still not good, measures 5,56V. Think it's broken, i've checked the writing on it and it's indeed a 4v3.
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

Re: No boiler communication

Post by AlfredB »

After almost throwing the otgw out of the window I now resumed working on it again. Replaced D11 and now I have much better values.

OK1 pin 8 => 4.17V
Cathode of D10 => 4.17V
Cathode of D9 => 1.98V
Both sides of R1 => 4.18V | 4.84V
OK1 pin 1 => 1.2V
OK1 pin 2 => 0V
IC1 pin 2 => 4.94V

If you connect pin 2 of the IC1 socket to VDD (e.g. pin 14), the voltage on X2 should drop to between 5V and 7V =>
Voltage now finally drops to 5,55V.

Finally one step further :-)

Still when I run test 4 I get the error "interfaces do not appear to be looped"

The next check with me now also fails:

Measure the voltage on pin 18 of the IC1 socket with pin 3 left open and again when it is connected to ground. In the former case the voltage should be below 1.25V (it is actually expected to be 0V). In the latter case the voltage must be above 1.25V (and will probably be close to 5V).

When I measure I get 0,79V on IC1 pin 18 (with IC1 pin 3 connected to ground = closed).

Volt measurement between IC1:18 and IC1:5 – 0V when 3 and 5 are open.
Volt measurement between IC1:18 and X1:0 – 0V when 3 and 5 are open.
Volt measurement between IC1:18 and IC1:14 – -0,71V when 3 and 5 are open.

Volt measurement between IC1:18 and IC1:5 – 0,79V when 3 and 5 are closed.
Volt measurement between IC1:18 and X1:0 – 0,79V when 3 and 5 are closed.
Volt measurement between IC1:18 and IC1:14 – 4,15V when 3 and 5 are closed.

Meanwhile I checked the values of all resistors again (measured ohm values) and they are all as specified. Also replaced the Q1,Q2,Q3,Q4 and Q5 to be sure they are all right. But this does not make any difference.

Anyone an idea?
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: No boiler communication

Post by hvxl »

AlfredB wrote:Volt measurement between IC1:18 and IC1:5 – 0V when 3 and 5 are open.
Volt measurement between IC1:18 and IC1:14 – -0,71V when 3 and 5 are open.
IC1:5 and IC1:14 are the power connections of IC1. The above two measurements basically boil down to a supply voltage of 0.71V. That's not good.
Schelte
AlfredB
Starting Member
Starting Member
Posts: 17
Joined: Fri Nov 18, 2016 2:28 pm

Re: No boiler communication

Post by AlfredB »

Found it finally! Replaced the two other zeners first. The last issue was caused by the optocoupler. The voltage of the test in the previous post never got above 0,91V. With the other opto (A827) it's now 1,27V (with IC1 Pin3 connected to ground). Now test 4 of the diagnostics firmware succeeds and reads:

OK1A high-to-low: 12us
OK1A low-to-high: 31us
OK1B high-to-low: 4us
OK1B low-to-high: 17us

Test number 5 now also returns a boiler voltage:

Enter test number: 5
Thermostat: 2.11
Boiler: 1.23, 1.25
Reference: 1.21

Thanks again for all support!
Post Reply

Return to “Opentherm Gateway Forum”