scanned register ranges

[lawther]

I wrote a quick script to scan the register file - issued 0x38 cmds to read a register/slot from 0x00/0x00 to 0xFF/0x0F and listened for a response. I ran it in a loop overnight to see what would happen.

Couple of things I discovered:

• register 0x3A slot 0x01 responds with 0x19. No idea what it means.
• for channels and light zones (type, name, state, multicolor, color, active), we've already documented a first block of 8 registers/slots. These appear to extend with a second block of 8 (ie a total of 16 channels/lights). In my pool, I have 3 actual lights, all are correctly set as 'blue'. All 16 registers in 0xD0-0xDF respond with 0x05 ('blue').
• with valve labels, we've documented 2 (0xD0-0xD1), but this seems to extend to 4 (0xD2 and 0xD3 respond). I have two valves, so I correctly get nul-terminated strings for the first two, and zeroes for the second two.

[marklynch]

Love it, good idea.

Interested to see what my controller returns for 0x3A slot 0x01 when I get back in a couple of weeks.

With the higher registers definitely should document and add them, but I suspect they are unused on all implementations, unless there is some way of controlling multiple connect 10s with a single touchscreen or there is a connect 10/16

Cheers
Mark

[zagnuts]

or there is a connect 10/16

I think there was a Connect 8 expansion module, so maybe that could be hooked in somehow and it would not surprise me if there's some legacy code crawling around there for backwards compatibility but I think the Connect 10 basically replaced the Connect 8 and is stand alone ie if you lived at the Palace of Versailles and needed to control more pools you could get more units, but they would run on separate networks and not know about each other directly so instead you'd probably be looking at products aimed at large commercial environments that could sync the fountains and lights up with Vivaldi. ;-)

[marklynch]

@lawther I'm not getting any response on the 0x3A slot 0x01 which is curious.

Could you share the script you ran on it?

Also I think it would be good if you can raise a PR with your branch - as I think adding that to the protocol would be really useful.

[lawther]

I got some more time, so I tried again to be sure. I typoed above, I issue cmd 0x39 (not 0x38) to request a register, and listen to 0x38 for the response. I can consistently repro this response:

    Request (cmd 0x39) register 0x3A slot 0x01:
    02 00 F0 FF FF 80 00 39 0E B7 3A 01 3B 03

    Response (cmd 0x38) returned 0x3A 0x01 = 0x19:
    02 00 50 FF FF 80 00 38 0F 17 3A 01 19 54 03

I've uploaded the script on a branch here https://github.com/lawther/pool-controller-code/tree/docs/extended-register-ranges/tools. It's pretty rough, not really intended for production.

The script runs in phases of actively sweeping the range, double checking, and passively listening to broadcasts. It runs indefinitely, building up evidence of what it hears, so if it hears multiple values for registers it'll record that.

If you're interested in keeping it, I will at minimum do a cleanup pass to bring it up to my normal python release state.

[marklynch]

Ok so the 3A 01 slot - is Solar temperature default. Which led me down a path of pretending to configure solar on my system which let to an initial implementation of this. Don't have solar and until I find someone who does leaving this as log only.
#73

[lawther]

Good find! My system did have solar heating back when it was first installed, but it had been removed before I got here.