Quinta 65

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

Moderator: hvxl

old_fart
Starting Member
Starting Member
Posts: 14
Joined: Wed Mar 30, 2022 11:20 pm

Quinta 65

Post by old_fart »

I have two Remeha Quinta 65 boilers. Are those compatible with the Opentherm Gateway? Additionally, since I have two, how do I distinguish between the two if I just have one gateway? Obviously, I can also buy two gateways, but how do you "pair" the two devices?

Can someone explain the architecture to me?

Right now, it's connected to an old PLC, but I'd like to disconnect that PLC and use a little bit more modern technology, such that it becomes possible to hopefully have more control over the device.

(I am a skilled programmer and I have programmed a bunch of ESP32s and Arduinos. )
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: Quinta 65

Post by hvxl »

The OTGW can work with boilers and thermostats that use the Opentherm protocol. Some features may not be available, depending on the level of support provided by the boiler and/or thermostat. The manual of the Remeha Quinta 65 specifies that it supports the Opentherm protocol. That means that it should be compatible with the OTGW.

The opentherm protocol is a point-to-point protocol, where one endpoint acts as the master (typically a thermostat) and the other endpoint acts as a slave (typically a boiler). When an OTGW is placed between the thermostat and boiler, there are actually two opentherm connections: Thermostat to OTGW, where the OTGW acts as the slave, and OTGW to boiler, where the OTGW acts as the master.

The OTGW has one interface of each type. So you can connect at most one boiler and one thermostat. If you want to utilize the OTGW to control two boilers, you will need two devices. These two gateways operate independently, unless you write/obtain special software to modify their behavior.

For more details on opentherm, I suggest you study the opentherm protocol pdf. Version 2.2 can easily be found on the internet. This is an old version, but most of the information presented there still holds true in later versions.
Schelte
old_fart
Starting Member
Starting Member
Posts: 14
Joined: Wed Mar 30, 2022 11:20 pm

Re: Quinta 65

Post by old_fart »

I have the NodoShop version running, but it's seemingly impossible to show any form of communication between the boiler and the gateway. Communication from a computer to the gateway and back is working fine and in Home Assistant (2022.8.7 with Frontend 20220802.0) I see about 175 disabled entities. I can enable those, but something like boiler temperature returns 0.0 degrees Celsius instead of something like 30. Basically everything has an uninitialized value suggesting no communication between boiler and gateway.

All I want is that I can set or read a value from the boiler and at that point I can continue with this project.

Is it possible to do the wiring from the two terminals labeled "boiler" to the terminals (terminal 1 and 2 of a 24 pole terminal block according to the Quinta 65 manual) wrong?

If I could just get the Home Assistant integration (https://www.home-assistant.io/integrati ... ntherm_gw/) to have a measurable effect on the boiler (or even that I could read some data as a first step, that would get me somewhere). Communication from the gateway to the PC is working, because for example the version of the firmware is being reported to Home Assistant.

Below is a log visible from OT Monitor running on Windows 7 resulting from pressing the "enable" button on the Home Assistant integration after it was previously in a "disabled" state:

Code: Select all

13:31:05.916999 PS: 0

13:31:05.962066 PR: A=OpenTherm Gateway 6.1
13:31:06.003226 PR: B=12:57 31-05-2022
13:31:06.030799 PR: C=4 MHz
13:31:06.054882 PR: W=A
13:31:06.079815 PR: G=00
13:31:06.111003 R00000000 Read-Data Status: 00000000 00000000
13:31:06.112902 PR: I=11
13:31:06.140951 PR: L=FXOMPC
13:31:06.165004 PR: M=G
13:31:06.188853 PR: Q=P
13:31:06.222083 PR: S=16.00
13:31:06.250889 PR: O=c21.00
13:31:06.283056 PR: P=Low power
13:31:06.306953 PR: D=O
13:31:06.332940 PR: R=D
13:31:06.358036 PR: T=11
13:31:06.381955 PR: V=5
13:31:06.586893 PS: 1
13:31:06.589238 00000000/00000000,0.00,00000000/00000000,0.00,0.00,0.00,0/0,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0,0/0,0/0,0.00,0.00,00000000/00000000,0,0,0,0,0,0,0,0,0,0
13:31:06.623161 PS: 0
13:31:06.728407 PR: I=11
13:31:07.102382 R00000000 Read-Data Status: 00000000 00000000
13:31:08.101160 R00000000 Read-Data Status: 00000000 00000000
13:31:09.102431 R00000000 Read-Data Status: 00000000 00000000
13:31:10.102462 R00000000 Read-Data Status: 00000000 00000000
13:31:11.106925 R00000000 Read-Data Status: 00000000 00000000
13:31:12.106125 R00000000 Read-Data Status: 00000000 00000000
13:31:13.105246 R00000000 Read-Data Status: 00000000 00000000
13:31:14.110297 R00000000 Read-Data Status: 00000000 00000000
13:31:15.107998 R00000000 Read-Data Status: 00000000 00000000
13:31:16.113150 R00000000 Read-Data Status: 00000000 00000000
13:31:16.759894 PR: I=11
13:31:17.108256 R00000000 Read-Data Status: 00000000 00000000
13:31:18.111105 R00000000 Read-Data Status: 00000000 00000000
13:31:19.109496 R00000000 Read-Data Status: 00000000 00000000
13:31:20.109909 R00000000 Read-Data Status: 00000000 00000000
13:31:21.112929 R00000000 Read-Data Status: 00000000 00000000
13:31:22.118008 R00000000 Read-Data Status: 00000000 00000000
13:31:23.115492 R00000000 Read-Data Status: 00000000 00000000
13:31:24.120550 R00000000 Read-Data Status: 00000000 00000000
13:31:25.118112 R00000000 Read-Data Status: 00000000 00000000
13:31:26.117143 R00000000 Read-Data Status: 00000000 00000000
13:31:26.786852 PR: I=11
<the pattern showing R00000000 Read-Data Status: 00000000 00000000 repeatedly, repeats from here on without extra information>
There is a line saying:

Code: Select all

13:31:06.283056 PR: P=Low power
Is that indicative of a problem?

The device was soldered by Nodoshop and they claimed that they tested it before shipping (which I don't doubt).

I do not have a physical thermostat connected, because I had kind of expected to just be able to set the dial to e.g. 30 degrees Celsius in e.g. the Home Assistant app and then see the boiler respond to that (which doesn't happen).

So, in short, I am completely stuck and need help. Perhaps it's just as simple as to tell the boiler to "use OpenTherm". According to the Quinta 65 manual, one needs to press the button with two arrrows vertically separated by each other for two seconds to test communication with the "controller" ("regelaar") and it's supposed to flicker then (which it does), indicating that communication is working, but I could well be wrong about this, since this is the first boiler I have ever tried to connect to OpenTherm Gateway.

I have been able to connect to it via telnet too (I disabled the Home Assistant integration while doing that).
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: Quinta 65

Post by hvxl »

People are capable of the most surprising things. So I can't rule out that the wiring can be done wrong. But it doesn't matter which way around the wires of the connector labeled "Boiler" are connected to the boiler. You can use a volt meter to check if there is any voltage on the terminals on the OTGW side. Expect mostly around 6V. The value will fluctuate when OT messages are being transmitted.

The "PR: P=Low power" report is just the response to a "PR=P" command, apparently issued by Home Assistant. So that is no indication of a problem.

I don't know what Home Assistant exactly does, but I suspect that the dial corresponds to the TT or TC command. Those commands only work with a compatible thermostat. But the first step is to have communication between the OTGW and the boiler, which you currently do not have. You should see BC0000000 responses to the R00000000 messages.

Since the OT test on the boiler indicates success, it seems that the wiring must be fine. A simple thing you can try first is to power down the boiler. Then, with the OTGW connected and powered, restore the power to the boiler. Perhaps it defaulted to on/off mode when it didn't see OT messages after power up the previous time. Also check the interface selection (parameter 9) is set to "internal Opentherm interface" (00). That's the default, so it probably is. But just make sure.

If that doesn't solve the problem, I suggest to go through the trouble shooting instructions (at least the parts you can do without a thermostat), to make sure the OTGW is working correctly.
Schelte
old_fart
Starting Member
Starting Member
Posts: 14
Joined: Wed Mar 30, 2022 11:20 pm

Re: Quinta 65

Post by old_fart »

hvxl wrote: Mon Sep 12, 2022 6:12 pm People are capable of the most surprising things. So I can't rule out that the wiring can be done wrong. But it doesn't matter which way around the wires of the connector labeled "Boiler" are connected to the boiler. You can use a volt meter to check if there is any voltage on the terminals on the OTGW side. Expect mostly around 6V. The value will fluctuate when OT messages are being transmitted.
The measurements have been done and I do recall that they fluctuated, but the value I do not recall. If the power supply is not good for some reason, would that be reported by the device? (That doesn't seem like an extremely advanced feature and lots of other types of devices have such features. )
You should see BC0000000 responses to the R00000000 messages.
Agreed and that definitely is not working.
Since the OT test on the boiler indicates success, it seems that the wiring must be fine. A simple thing you can try first is to power down the boiler. Then, with the OTGW connected and powered, restore the power to the boiler. Perhaps it defaulted to on/off mode when it didn't see OT messages after power up the previous time. Also check the interface selection (parameter 9) is set to "internal Opentherm interface" (00). That's the default, so it probably is. But just make sure.
The suggestions you made have been tried and there was no difference in output on the serial console. The manual for the Remeha Quinta 65 is completely awful w.r.t. service level access, etc., but we managed to find parameter 9 and it was set to 00.
If that doesn't solve the problem, I suggest to go through the trouble shooting instructions (at least the parts you can do without a thermostat), to make sure the OTGW is working correctly.
There isn't really any evidence that the Quinta 65 has ever worked, right? The manual is shared with the Quinta 45, which is listed as working.

I thought there were differences on the Nodoshop version of your board. It seems impossible to then follow your instructions, since the board layout might be different, etc.
If the thermostat is still unable to communicate with the boiler in this mode, you should focus your debugging efforts on the hardware of the opentherm line drivers. Test #4 of the diagnostic firmware may be of help with that.
This suggests it's the boiler hardware, if I understand you correctly, but debugging someone's proprietary hardware is going to be even a bigger issue.

It's somewhat late, but it seems to be a project with a scope beyond what I had imagined it to be. I guess one of the things we would have to know is whether the boiler is communicating information or the opentherm gateway at any given point in time. I suppose the professional way to do that is to build a FPGA that can listen to the wave and tap into that signal, but that would be another two years of my life before I get something to work, which sounds like a horrible plan.

The boiler I am trying to connect to was part of a hugely overdimensioned cascade system. Is it possible that it's configured as a slave and that it has to be a master for example or something similar? (I would have expected the manual to say something about such conditions, so I guess not.)

I find selecting a reasonable next step to take extremely difficult. Should I ask Nodo-Shop to send another device for testing? Should I contact Remeha and ask them? I think both aren't great steps. Would anyone find it interesting to have physical access to a Quinta 65 device (we have two, as said before) and obtain the log files (as on your website), such that the "value" of the gateway becomes larger (due to more supported devices)?
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: Quinta 65

Post by hvxl »

old_fart wrote: Mon Sep 12, 2022 11:12 pm If the power supply is not good for some reason, would that be reported by the device? (That doesn't seem like an extremely advanced feature and lots of other types of devices have such features. )
The OTGW from the Nodo-shop has LEDs for the different supply voltages. These LEDs should all be steady. If you see them fluctuating, or not lit at all, then there is probably an issue with the power supply.

If you believe it is so easy to report power supply issues, I'll happily accept your patches to implement that feature.
Schelte
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: Quinta 65

Post by hvxl »

I would definitely expect the OTGW to work with a Quinta 65.

Not all of the trouble shoot instructions are applicable to the OTGW from the Nodo shop, like the power supply and serial interface parts. Just skip those parts. But most of the remaining tests refer to PIC pins and the thermostat/boiler connectors. With a little imagination those test can still be performed.
Schelte
old_fart
Starting Member
Starting Member
Posts: 14
Joined: Wed Mar 30, 2022 11:20 pm

Re: Quinta 65

Post by old_fart »

hvxl wrote: Thu Sep 15, 2022 12:29 pm
old_fart wrote: Mon Sep 12, 2022 11:12 pm If the power supply is not good for some reason, would that be reported by the device? (That doesn't seem like an extremely advanced feature and lots of other types of devices have such features. )
The OTGW from the Nodo-shop has LEDs for the different supply voltages. These LEDs should all be steady. If you see them fluctuating, or not lit at all, then there is probably an issue with the power supply.

If you believe it is so easy to report power supply issues, I'll happily accept your patches to implement that feature.
I remember one green led continuously on and another green led blinking about every second. Both were located near the back of the board (which is where the ethernet connector is located) on a strip of leds (no red leds were visible anywhere on the board).
old_fart
Starting Member
Starting Member
Posts: 14
Joined: Wed Mar 30, 2022 11:20 pm

Re: Quinta 65

Post by old_fart »

hvxl wrote: Thu Sep 15, 2022 12:29 pm
old_fart wrote: Mon Sep 12, 2022 11:12 pm If the power supply is not good for some reason, would that be reported by the device? (That doesn't seem like an extremely advanced feature and lots of other types of devices have such features. )
If you believe it is so easy to report power supply issues, I'll happily accept your patches to implement that feature.
Conceptually, it seems easy, but it might be impossible or hard to do it economically.

For example, ultracapacitors that would allow the device to work with even the most horrible power supply. If a true quality is required, then there is probably a way to integrate (as in the mathematical concept) the deviation from a theoretically perfect power signal and display that as a number, unless turned off in configuration. I would expect almost every electronics product to have something like that in the development version of the product.

Apparently, that's called a low-frequency spectrum analyzer and one can impement that using a 89C52 (link to a design: https://www.iasj.net/iasj/pdf/004f2e517ddada93).

Now, I understand there's a difference between pointing at some design and making an actual electronic design that would fit this particular application, but it seems like just an engineering thing, not a science endeavour (which in a loose sense makes it "not extremely advanced").

I think the economic value (which is a proxy to the value to society) of the OpenTherm gateway is related to its support costs. Another way to get around that is to have a cheap test kit available such that you can fake a range of boilers and thermostats.

Please let me know if I told you something of interest, but I doubt there's anything I could have thought of that you, as the unquestionable expert, haven't. I am sorry for not being able to offer you content that's of greater interest.
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: Quinta 65

Post by hvxl »

Someone else had a similar problem. Fortunately he has a thermostat, so he could do more extensive tests with the diagnostic firmware.

It turned out that he could get things working by lowering the reference voltage setting. You may want to try that too. Since you don't intend to use a thermostat with the OTGW, you can go all the way down to 0, if necessary.
Schelte
old_fart
Starting Member
Starting Member
Posts: 14
Joined: Wed Mar 30, 2022 11:20 pm

Re: Quinta 65

Post by old_fart »

VR=0 has no effect on the boiler sending data including after a reset. (Other than VR: 0). I wrote a program to easily set such values independent of the network the OTGW is connected to.

When connecting to the thermostat with VR=4, which I think sends opentherm signals, I just get (via serial):
R80000100
R80000100
R80000100
R80000100
which is the same as reported here: https://www.domoticz.com/forum/viewtopic.php?t=11643 in 2016. I have no idea what the meaning is of "R80000100".

I also get the message "Low power" via serial, even though the power supply has a total rated supply of 3x1A and I have never ever had such issues with any hardware ever before. I have read various messages about measuring with a multi-meter between various contacts, but IIRC those instructions were not that easy to follow. I'd expect very high resolution images, such that it is completely obvious which contacts are those of concern and whether or not a boiler needs to be attached (which would highly complicate matters, because boiler rooms are typically small and lengths of wires are also short, etc.) when the test is being done, etc.
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: Quinta 65

Post by hvxl »

There are 3 power levels associated with Smart Power: Low power, Medium power, and High power. Low power is the default level. The report is in no way related to the power supply used to power the OTGW.
Schelte
old_fart
Starting Member
Starting Member
Posts: 14
Joined: Wed Mar 30, 2022 11:20 pm

Re: Quinta 65

Post by old_fart »

I verified that it does work with another OpenTherm boiler of a different brand. Does that just mean that the OpenTherm gateway is not compatible with the Quinta 65, even though it is claimed that the Quinta 65 has the same board as a Quinta 45 (which is listed as being compatible) or does that just mean that there was a user error while connecting it last time? I don't really have any particular reason to assume anything being different. I completely expect that if we were to try it another time, it will just not do anything useful again.

I think the assertion "Quinta 45 works -> Quinta 65 works" is wrong myself, but I have been wrong before.

Should I call Remeha for an explanation? It seems a bit of a waste of money to have two boilers with a new price of 3,000 euros that are essentially worthless without being able to control them via the opentherm gateway.
hvxl
Senior Member
Senior Member
Posts: 1965
Joined: Sat Jun 05, 2010 11:59 am
Contact:

Re: Quinta 65

Post by hvxl »

While you have been quite unreasonable with your expectations for an open-source project, I have created a new version of the diagnostic firmware that may be helpful for you: 1.3 for PIC16F88 and 2.1 for PIC16F1847. These versions are more capable of performing tests without the presence of a thermostat. For tests #3 and #5, the OTGW will now generate test messages toward the boiler.

After loading the diagnostic firmware, you can run test #5 to determine which reference voltage setting to use. After setting that, you can run test #3 to verify if you get any responses from the boiler and if so, what they look like.

By the way, there is no list of compatible equipment. If you are referring to the equipment matrix, those are just thermostats and boilers for which users have uploaded log files. There is no way for me to verify if the make and model information provided by those users is correct. I have to rely on the goodwill of the users who upload a log file. As the page says: "I obviously cannot give any guarantees on the information provided below. It's a best effort attempt to provide some more details on Opentherm equipment than are readily available from other sources. However, I can not be held responsible if information in these tables turns out to be less than accurate. "
Schelte
old_fart
Starting Member
Starting Member
Posts: 14
Joined: Wed Mar 30, 2022 11:20 pm

Re: Quinta 65

Post by old_fart »

I think it's more likely that the OTGW is not able to fake the presence of a thermostat of a given brand, which for some boilers probably makes a difference.

I think your equipment matrix is not precise enough, because it doesn't record whether or not a thermostat (and if so, which one) was connected when the data was received. This makes the data almost useless, unless you claim that OTGW is a perfect simulator of an OpenTherm thermostat (in which case, I wouldn't even have to buy an OpenTherm thermostat, which is something I have already planned on doing, since finding out that it likely does make a difference).

> While you have been quite unreasonable with your expectations for an open-source project

I just expect that it speaks OpenTherm and that the commercial project also speaks OpenTherm (which it does according to the manual). I think you should just tell people that _just_ having a boiler without a thermostat connected is not at all expected to work. You recommend people not to use a software thermostat for various reliability reasons, but that's a _completely_ different aspect.

I think the feature for performing tests to figure out the voltage, should be part of the production image, potentially with a tiny chip to store a working value in some persistent small electronic memory, which is what I explained months ago already. I have the impression that test #5 is derived from my idea. Now, you could just have said "Thanks, that was a good idea. I implemented a slightly simpler version of this. ". I don't see the need for your pedantic tone. What is unreasonable about my expectations? In fact, I did buy a commercial product (the one from the Nodo shop). It just happens to not ship with code that works for my use case despite people telling me that it should work... In short, the advice given was, while perhaps of an expert level, not sufficient for the given circumstances. How is that my fault?

I think that the quality of an open-source project should not be less than that of a commercial project and indeed the Linux kernel is objectively better than the Windows kernel for various applications (for example, large parts of HPC). Why do you think an open-source project should not be held to the same standard?

It's not clear to me why the default firmware doesn't automatically do test #5 at start. Who cares about that one second? That is, the code should just check whether the right chip is there, do the test, and then set the voltage. Really, why would you ever want to have different images? In fact, it would make much more sense, if you just had the compilation structure in some build system where people can mix and match the various features. You could have a wizard where people can just say which features they want in an image and then those are compiled in. Any Autotools based system would be able to do that just fine.

This software could be improved with code similar to what is in nmap to identify remote machine types depending on the amount of system memory you still have available. A feature that would be able to load code off a remote server would also be useful.
Post Reply

Return to “Opentherm Gateway Forum”