Skip to main content

You are here

New Boiler / central heating system with opentherm

35 posts / 0 new
Last post
Jeroen
New Boiler / central heating system with opentherm

Hi, My boiler / central heating systems is up for renewal. I like to buy a new system, and this new system should be able to communicate with my home automation system (jeenodes). Looks like Remeha has the best interface possibility. (I am looking at the Remeha Calenta CW5 Combi Comfort System)

Or are there better alternatives?

regards jeroen

amvv

I have managed to get a JeeNode talking to a Remeha Calenta (35 something) using an opentherm interface I designed and built myself! I put some info somewhere in the wiki. It has been running for almost 2 months now with no problems! So I would say that any boiler supporting OpenTherm should be OK!

Jeroen

Thanks !

Jeroen

Hi, My "old"boiler replaced with a Remeha Calenta, somehow this boiler popped up as the best central heating system and just by coincidence it is OpenTherm based ! ;-) uhhh I found this document on the wiki (http://jeelabs.net/projects/cafe/wiki/OpenTherm_interface) (thanks alex). and found on the internet also this approach (http://blog.hekkers.net/tag/remeha/). The first one based on opentherm (directly) and second one by using the recom interface. Not sure if it is possible to connect something to the opentherm interface if there is also a thermostat connected to the boiler. (isense) My favorite approach would be the one with the least chance to get a cold shower in the morning. So far without testing the second option look like the less dangerous one if I only read information from the boiler, using one of the options, http://www.kmitl.ac.th/~kswichit/ap275/ap275.htm (only the top tx part) or http://blog.hekkers.net/wp-content/uploads/2010/10/db25sch.png. (max32)

(edited: for testing the USB-bub shoul be ok I think (http://www.barrydegraaff.tk/blog/index.php?blog=20120323093423) any comments on this ? Or better solutions ? Regards, Jeroen

Jeroen

Last days I created a small Opentherm interface (based what I found on the Internet). This for only lurking the data from the boiler and ... Bingo. Here is the first part of the remeha's "banga lijst"

0 6F 7F FF 83 FF FF 7F 17 E0 9B 93 3F E8 BF FF 80 19 3F FF 0 5 0 0 80 0 1 1F 90 FF 67 19 0 11 1 FF 83 F2 80 0 0 F2 7C 7E 80 7 FC 80 90 1 31 CC 0 11 0 0 3F F3 7F FF 91 F8 CD E0 80 0 7F 7F 90 1 31 C9 0 11 0 0 80 19 0 0 90 10 14 80

So far I have no idea what these bytes represent but it feels already good.

Robert

Hi Jeroen, the Opentherm frame format can be found in the Opentherm Specification, which can be found here, including a list of the standard Data ID's. I'm curious whether you'll also see frames with a OEM Data ID about once a minute.

Jeroen

Today I connected a usb-bub to the Calenta (ground tx and rx) downloaded the software from the remeha website and got results. So now I have two options to read information from the boiler.

AttachmentSize
recom.jpg293.26 KB
Jeroen

After JCW's daily Opentherm updates, I started my Opentherm interface again. My Jeenode sends OT bytes to the server. Most OT frames are fine, but there are also a lot data-id's with no data value. (Example 0 18 0 0 (CH water pressure, or 128 109 0 0 ) My thermostat shows 1.6 bar) so it must be send from the boiler to the thermostat. So some of the frames never contain data and most contain always data. any ideas? Does anyone know the meaning of data-id 109,110,111 and 112?

jcw

I would expect that a thermostat sends out all sorts of requests, but only some of them will get an answer - depending on the boiler model?

Jeroen

I agree, but that does not explain the OT message "0 18 0 0" (CH water pressure) It does not contain any value(last two decimals) but on the display of my thermostat and boiler there is displayed 1.6 bar. So lots of fun the next few days.

Mars

Weird problem, Jeroen. Are you sure it's not a hardware failure causing you to read zero's??

I'm still hoping that the OpenTherm gateway jcw is experimenting with will work out, as this makes it possible to also set the setpoint, etc.

jcw

Here are the packet counts I see over the past couple of days:

logger/20121122.txt 2183 logger/20121123.txt 3742 logger/20121124.txt 5186 logger/20121125.txt 3162 logger/20121126.txt 958 logger/20121127.txt 2046 logger/20121128.txt 1564 logger/20121129.txt 3999 logger/20121130.txt 3644 logger/20121201.txt 1797 logger/20121202.txt 4253 logger/20121203.txt 3867 logger/20121204.txt 3231

No manual resets in this period, I think. Note that only changes and periodic 4-minute repeats are sent, so variation is normal. Couple of days are a bit low, but it always recovers. I'll need to dig deeper to see whether there are long periods of off-time, and what the relation to heat demand is. Most of the time this thing runs on the thermostat, with occasional manual overrides to bump it up a notch for the evening.

I could produce a log extract of all messages from say December so far, if anyone wants to look into it. They're just a bunch of timestamped JeeLink output lines, saved to file.

Mars

These are the logfiles using the OpenTherm gateway connected at the boiler side?

Donno if I have time to really look at them this week, but you may post them...

jcw

No - incoming packets. OTG is tied to a JeeNode, there is no galvanic connection from OTG to anything.

Mars

I was a bit cryptic I guess...

You mentioned bedore that there was a problem with resets if the OTG was connected as boiler - OTG - 10m cable - thermostat as opposed to the OTG at the thermostat side, ie boiler - 10m cable - OTG - thermostat.

I did read your post where you connected the JeeNode to the OTG collecting and sending the messages.

If I would build the OTG, it would be placed next to the boiler, so also with about 10m cable to the thermostat (iSense). So I'm looking for a working solution, and therefore are curious about your results.

Jeroen

Could be a HW or SW problem. (Anyone a working sketch with rfm12b sending opentherm packets?) I'll change the hw this weekend for something else.

MichelV

Got a new central heating system as well a few weeks ago, the Remeha Calenta something. Seems to be a popular one here ;-).

Jeroen

Still not sure how to go forward from this point. Look like the cicuit based on http://jaagpad.blogspot.nl/2011/09/opentherm-circuit-diagram.html and this sketch :http://www.hekkers.net/files/OT_decoder.rar works best. This cirquit:http://palebluedot.nl/jml/projects/arduino/24-openthermmon.html gives often connection problems between the boiler and the thermostat. The sketch on the wiki at jeelabs: http://jeelabs.net/projects/cafe/wiki/OpenTherm_interface gives me only errors (start-stop bit error and bad data). I also have at least 10 meters of wire between the boiler and opentherm circuit, so perhaps the same problems as described in post http://jeelabs.org/2012/11/25/bi-directional-signalling-and-power/. any hints / ideas ? regards Jeroen

Mars

Weird. You have problems with 10m cable from the boiler, and jcw not, bit with a different gateway...

It seems that those circuits don't like long cables, as opposed to the OT spec of max 50m cable.

What is so different in the boiler and / or thermostat that all those gateways don't have???

martynj

@Jeroen, the various circuits treat the wiring pair from boiler to thermostat with great caution, isolating it from the logic side through optocouplers. Caution is indeed needed, in rare cases a pair could carry mains potential - but in general, it is intrinsically safe, limited to a low current flow and 42Vmax between the wires. The optocouplers allow the loop to "float" with respect to the ground potential of the detector/gateway.

Here is a source of some confusion. This isolation is good to avoid ground loops (where the earth potential of master, slave or detector/gateway may have different definitions of 'earth' - especially under fault conditions). The loop is "floating" w.r.t. the local ground, breaking any ground loop.

This does not mean that the loop should be without a ground reference at one and only one point. Completely floating means that the actual potential to ground gets defined by some complicated mixture of stray capacitance, leakage resistances and RFI, varying over time.

JCW wrote about this effect - unable to even touch the loop with a scope probe without large disturbances. Following this recipe should avoid that.

  • Check that the slave side (i.e the boiler) has a good ground reference. Safety requires that the metal casing and pipework are securely bonded to a trusted ground reference. Many local wiring codes insist on a minimum wire gauge, colour code and safety labels "Do Not Remove". If you are in any doubt about the integrity of this earth path - call a qualified electrician

  • Check that one side of the loop has one and only one ground reference. One side of the pair should show zero volts on AC, then DC with respect to the ground you have just verified above. This reference is usually made by a simple link at or near the boiler terminal block. The boiler schematic may be provided, else inspect visually.

  • Check that the master (i.e. thermostat installation) allows the pair to float (e.g. no link to a local ground - fully encased plastic designs do not need a ground connection, though there may be a terminal block position for one that goes nowhere internally)

Sounds like a lot of checking - you may be thinking "How come my existing setup worked fine and doesn't pass the checklist?" Well, the legacy wiring may date from an installation with a mechanical thermostat that just didn't care about fancy "floating" explanations - ON or OFF, an open or shut case ;-)

As always, take great care near mains-powered circuits. SAFE is SWITCH OFF.
Only turkeys should get roasted at Christmas!

Jeroen

Martin, first of all thanks for the reply. I did a lot of checking today. Boiler is connected to earth ground. (so does the plumbing) One side of the Opentherm circuit was already connected to earth ground (0 volt between one wire of the OT circuit and earth ground. To be sure I connected a extra wire just outside the boiler. I moved the whole setup (OT circuit, laptop and thermostat)to the boiler and disconnected the long wires. All with the same result: some OT messages are always good some are always 0 what is always good: Control Setpoint, Room Setpoint and Room Temperature. Status switches between 0 and 1 (zero no heating or DHW) and 1 ch active) the rest of the bits are not present attached a opentherm tar file , just a few minutes (rename zip to tar)

AttachmentSize
ot.zip60 KB
martynj

@Jeroen, well that elimates "floating"/noise pickup issues. Leaves the interface circuit itself and/or the software...

Does your current sketch check for badly formed frames?

Jeroen

Hmmm after a good glass of wine, I ripped everything from the breadboard, and started again soldered everything again with fresh components, and yes Yes YES Yabadabadoo.!!!!!!!!!!! All kind of new information !

Mars

Great. The power of a glass of wine is largely underestimated by most ppl !!

martynj

@Jeroen good, perhaps the Shop should stock that wine next to the Hopje as a debugging aid!

When you get around to posting the new results, could you add a 'raw packet' sequence so it is easier to see how they interleave in time?

tankslappa

Proposal seconded! (The wine one)

:-)

Mars

Proposal seconded! (The wine one)

:-)

Oh, noooooooo| The guy who likes to zap his neighbours TV found this thread! HIDE !!!

tankslappa

You can run, but you can never hide.

Jeroen

Martin the log with timestamps is in process. Perhaps also champagne to celebrate success !!! And like the hopjes a 9-pack. I think JCW is already creating a entry in the shop: http://jeelabs.com/products/xtra-dummy

martynj

Looking at the forecast weather - perhaps switch that to glühwein.

tankslappa

Mulled wine eh?

Works for me!

Then again, I'm off to Hungary for Christmas and New Year, so I imagine they'll have something far more lethal on offer!

Pages

Premium Drupal Themes by Adaptivethemes