[OpenTRV-interest] Tinkering kit / Conrad radiator valve and security code

[Damon Hart-Davis]

Hi,

On 19 May 2014, at 10:24, Alasdair Macdonald <EMAIL ADDRESS HIDDEN> wrote:

> Last night I tried to fit my Conrad TRV to a radiator, in order that I could try to establish a dialogue between the TRV and my OpenTRV tinkering kit.
> 
> Fitting the TRV proved to be a very frustrating experience, primarily because the Conrad instructions assume that you have additional kit, and that some relevant instructions will be available in the instructions for that kit.
> 
> Conrad instructions for "Valve Drive 57 00 56":
> 
> "In the case of valves made by Danfoss, you first have to put on one of the supplied adaptors (see the operating manual for the complete system)”.


Sorry that this has been so frustrating: I haven’t managed to fit one of these units to anything other than a vanilla “Danfoss” M30x1.5 fitting, so you may be our new “expert”…  %-|P

I’ll comment on some of the OpenTRV unit serial output below, since that’s all my fault...

> 
> Here's a link to the Conrad website for the Valve in question, which has been sitting next to my PC for the past 6 or 7 weeks:
> 
> http://www.conrad-electronic.co.uk/ce/en/product/570056/FS20-Indivdual-Wireless-Thermostat-Actuator
> 
> Manual:
> http://www.produktinfo.conrad.com/datenblaetter/550000-574999/570056-an-03-ml-Ventilantrieb_de_en_fr_nl.pdf
> 
> I found the clue that I required in the manual for the other part in this 2-piece set:
> 
> http://www.conrad-electronic.co.uk/ce/en/product/617500/FHZ-1000-Wireless-Control-Centre-Economy-Set
> 
> Here's the link to the manual for the FHZ1000 set:
> 
> http://www.produktinfo.conrad.com/datenblaetter/600000-624999/617500-an-01-en-SPAR_SET_FHZ_1000.pdf
> 
> The photos marked figures 10 / 11 / 12 are included in the instructions for the Valve, tiny pictures of course, but the photos marked figures 14 / 15 / 16 are NOT in the valve manual. The fittings on my radiators required the adaptor that is identified in figure 14, "type RA". It wasn't clear to me that this was the correct fitting, mostly because it's by no means a snug male/female fit. The nut and bolt hold the collar around the existing metalwork, but there is no lip or ridge that could prevent movement of the collar. I don't know if this is a peculiariaty of my fitting, or if all Danfoss "RA" would be like this (mine doesn't exactly resemble the photograph).
> 
> Additionally, I used a plastic extender for the radiator actuator pin; this was supplied with the Conrad Valve (and is probably visible in the manual in figure 15). But right now I can't say that it's correctly positioned, or if the movement in the valve actuator up to the time might have knocked it out of position. The pin and extender piece - both currently hidden from view - have rounded tips, so although the pin does have a recess that allowed me to place it appropriately on top of the existing radiator pin, I don't know if the movement of the actuator pin will cause it to move out of place.
> 
> Which brings me to my next point.
> 
> I followed the instructions in the Valve manual for setting up the valve. Refer back to 550000-574999/570056-an-03-ml-Ventilantrieb_de_en_fr_nl.pdf . As the valve is powered, I see the readouts described in the manual. First C1, then "93", then C2, then "74", then A1. I understand that 93 / 74 is the house code pre-programmed into this valve.
> 
> Manual states to press the button for 3+ seconds to put the valve into a state where it may expect to receive a "security code". The valve has been awaiting this security code since ~1am last night.

You should hear another ‘beep’.

> 
> I logged in to the openTRV controller (it has either lost ~10 minutes in the past 7 weeks, or I must have switched it off / on at an appropriate time -10 minutes past midnight?). I previously reported that my controller did NOT reboot when I fired up the Arduino IDE. Certainly today (a few minutes ago), it DID reboot; the first things that I saw in my terminal screen were the reboot messages.

Internally the OpenTRV unit does not keep time through a power-up or reset and simply stores it inside in 15-minute increments, so may at worst case jump forward or back that amount, though the typical error is much smaller.


> 
> Last night I tried to execute the House Code command to the controller. It looks as though it has succeeded somewhat; every line of output now terminates with "HC93 74 s".
> 
> I also got some new "FHT8V" messages, and I don't know what to make of these.
> 
> 
> 
> Here's some output from last night. I've removed some greater than signs from the output, in case your mail reader might misinterpret or hide the lines in question.
> 
> 
> "Normal" output, before I started tinkering:
> 
> =F0%@22C7;T0 50 W255 0 F255 0 W255 0 F255 0;S10 10 21 cffO


Frost mode, valve open 0%, temperature 22+7/16C; nominal time 00:50, no LEARNed programmes set; Setpoint currently 10C (10C frost, 21C warm), in ‘c'omfort mode, this and next hours predicted to be frost mode, highly probable ‘O’ccupancy.

> 
> 
> So I tried to send the "Set Wireless House Codes" command, per the help text, and using the codes displayed by my Conrad valve:
> 
> H 93 74
> =F0%@22C7;T0 50 W255 0 F255 0 W255 0 F255 0;S10 10 21 cffO;HC93 74 s
> 50:52:0 FHT8V SYNC...
> 
> So the output has ";HC93 74 s" appended, and some new output has begun to appear.

Yes, you set the house code and the trailing ’s’ in the section showing that indicates ’s’ynicing with the valve which takes a couple of minutes.
> 
> <snip>
> 
> More "normal" output, with the HC stuff at the end:
> 
> =F0%@22C8;T0 51 W255 0 F255 0 W255 0 F255 0;S10 10 21 cffO;HC93 74 s
> 
> 
> 
> And some exciting new output, which I don't know how to interpret (this was around 1am; the OpenTRV board seems to be saying that it is 0:55am in the output below):
> 
> 
> 52:56:73 FHT8V SYNC FINAL

The OpenTRV unit should now be in sync with the valve and the valve should have gone ‘beep’/


> 54:52:73 FHT8V TX

A transmission was made to the valve (happens about every 2 minutes).

> =F0%@22CA;T0 55 W255 0 F255 0 W255 0 F255 0;S10 10 21 cffO;HC93 74

The sync is over (no trailing ’s’ in the final section).

> 
> 
> 56:48:73 FHT8V TX
> 58:44:73 FHT8V TX

A couple more transmissions (with timestamps, basically minutes:seconds:s/128).

> =F0%@22CA;T0 58 W255 0 F255 0 W255 0 F255 0;S10 10 21 cffO;HC93 74
> 
> 
> 
> 
> Here I input the correct time at 1:04am:
> T 01 04
> =F0%@22CA;T1 4 W255 0 F255 0 W255 0 F255 0;S10 10 21 cffO;HC93 74
> 
> 
> 66:40:73 FHT8V TX
> 
> 
> 
> This morning I am getting stuff like:
> 
> 610:26:74 FHT8V SYNC FINAL
> 612:22:73 FHT8V TX
> =F0%@23CD;T10 12 W255 0 F255 0 W255 0 F255 0;S10 10 21 cffO;HC93 74

It should not SYNC again unless there is a problem or you power-cycle or reset it: if you’ve got a complete transcript of the output I’d like to see it.  Possibly the OpenTRV unit reset due to a bug.  There have been a number of updates to try to address this.

> 
> 614:18:73 FHT8V TX
> 
> However, my Conrad TRV still seems to be awaiting receipt of a "security code". I would like to send this from the OpenTRV controller, but I can't figure out from the help text what I should send (and is timing important??). The valve STILL shows "AC" with a flashing antenna; so according to the manual it's awaiting said code.

I think you’ve got the valve into a state that I’ve never seen, possibly a pairing state.  I haven’t written code to use any such feature yet.

Can you please go through out set-up guide starting with the batteries out of both the valve and OpenTRV units?

http://opentrv.org.uk/documents/

Rgds

Damon