[OpenTRV-dev] Dongle

[Robert May]

Thank you to everyone contributing to this thread.  I'm following with
interest and would like to help out (although I'm not quite there in terms
of having the time right now).    I have a set of Radbots waiting for me to
be able to implement a stats hub and boiler controller, and as my plans
involve a Pi anyway I had been hoping that there would be an easy way to
add a receiver to the Pi so I could combine it all.

I have a relatively complex mix of UFH and radiators and I'm planning on
building a combined controller, with the RadBots looking after the
radiators and a custom system to look after the UFH.   I'll write it up as
I go along (once I get started).  PLan is to have it operational for next
winter (but I would have said the same last winter!)

Rob.

On Sun, 10 Jan 2021 at 10:38, Tristan Keen <tristan.keen at gmail.com> wrote:

>   Thanks to all your help - though I'd already ordered an Adafruit
> RFM69HCW 868 MHz "Bonnet" for my Raspberry Pi (from Pimoroni), on the
> strength of a Hope datasheet suggesting it was an easy replacement part.
>
>   Unfortunately while the pinout might be the same, the registers are
> totally different so it's been slow going; especially when I realised
> Manchester encoding is implemented completely differently between the two -
> I was about to give up & buy a new module. Luckily, despite
> https://github.com/opentrv/OTRadioLink/blob/345b45f9df92eac78caa483d4230a11bde21d7d8/content/OTRadioLink/utility/OTRFM23BLink_OTRFM23BLink.cpp#L787 your
> (Christoph's) config actually shows Manchester was turned off.  Trying that
> as a last-gasp effort seems to work - I can now receive valid secure frames
> with sensible IDs matching some of the hex printed inside the Radbots, and
> incrementing message_counters from the trailer.
>
>   Now to read up on the encryption approach & get that working in Python,
> then I'll try reprogramming the key on one of the radbots to a known
> value.  Yes, my current code is similarly littered with odd notes, some
> copied in material and private IDs, I'll clean it up a bit & clone it
> publically when I've got my MVP working.
>
> Thanks,
> Tristan.
>
>
> On Thu, 7 Jan 2021 at 19:23, Damon Hart-Davis <dhd at exnet.com> wrote:
>
>> Purged of third-party copyright material, would you consider adding that
>> to an appropriate place in our Wiki or one of the GitHub repos?
>>
>> Good to have register settings for another radio widely available (and
>> the Pi stuff)!
>>
>> Note that there is one non-JSON data item in there, the valve %, which is
>> a single binary byte easy for a boiler controller with a small MCU to use...
>>
>> Rgds
>>
>> Damon
>>
>>
>> On 7 Jan 2021, at 19:14, christoph at winterstiger.at wrote:
>>
>> Hi Tristan,
>>
>> That’s right, I have this working. I played around with multiple
>> different modules and I think the cheapest option is a TI CC1101-based
>> module (lots of options for under GBP 10 on eBay). Most of them come with
>> antennas that should work fine for development purposes if you’re in the
>> same room as the RadBot.
>>
>> That module (and many similar ones) uses the SPI-bus for communication,
>> which is very easy to use on an RPI, and there are also options for other
>> platforms, e.g. CH431A-based modules (but I haven’t tried those yet). This
>> takes voltage + ground + 4 SPI wires + 1 interrupt wire (interrupts when
>> the module received a packet). On the RPI, no other hardware is required
>> and I think a CH431A wouldn’t need any other external components either.
>>
>> For my own experiments I wrote a monitoring app that lets me configure
>> the module on the fly so I can play around with all those RF-related
>> options, but I foolishly copied a bunch of copyrighted text from the
>> various datasheets into it, so now I’ll have to remove that before I can
>> publish it. I attached the RadBot config I use, which gives you all the
>> register settings for the CC1101. Once all the registers are set, you just
>> need an event handler for rising edges on the interrupt lines to get the
>> raw packets. Those are encrypted by the RadBot with a pre-set key, so they
>> need to be decrypted. Once that is done, the actual data decoding is
>> trivial as it’s just a simple JSON object.
>>
>> I’ll try to find some time over this or the next weekend to clean up my
>> code so I can publish it.
>>
>> Cheers,
>> Christoph
>>
>> *From:* OpenTRV-dev <opentrv-dev-bounces at lists.opentrv.org.uk> *On
>> Behalf Of *Tristan Keen
>> *Sent:* 30 December 2020 10:33
>> *To:* opentrv-dev at lists.opentrv.org.uk
>> *Subject:* Re: [OpenTRV-dev] Dongle
>>
>> Hi Christoph,
>>
>> I read from the mailing-list archives that you recently managed to "...
>> receive the messages on a raspberry pi with a cheapo 868Mhz transceiver and
>> it works fairly reliably".
>>
>> Would it be possible to share a bit more detail on what actual
>> transceiver you used with what Pi, plus any other hardware mods you had to
>> make? Additionally the code you used to decode the messages - even in
>> partial form if you prefer. I'm an experienced developer with some hardware
>> knowledge, so anything rough would still be gratefully received.
>>
>> I'm trying to make up something similar to the Stats Hub to collect
>> information broadcast by the OpenTRV/Radbots, but without building so much
>> custom hardware if possible...
>>
>> Thanks in advance,
>> Tristan Keen.
>>
>> <cc1101-radbot.cfg>_______________________________________________
>> OpenTRV-dev mailing list
>> OpenTRV-dev at lists.opentrv.org.uk
>> http://lists.opentrv.org.uk/listinfo/opentrv-dev
>>
>>
>> _______________________________________________
>> OpenTRV-dev mailing list
>> OpenTRV-dev at lists.opentrv.org.uk
>> http://lists.opentrv.org.uk/listinfo/opentrv-dev
>>
> _______________________________________________
> OpenTRV-dev mailing list
> OpenTRV-dev at lists.opentrv.org.uk
> http://lists.opentrv.org.uk/listinfo/opentrv-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opentrv.org.uk/pipermail/opentrv-dev/attachments/20210110/e46f1adb/attachment.html>