Thanks (plus an issue with Additional-attributes branch)

[washboy]

This is not a proper issue report (although I have been experienced difficulty with the Additional-attributes branch). It's a thank-you to @NigelCoxon.

I saw your invitation to @tomtokic, last November, to try your custom component. Tom had very kindly helped me get up and running with bretttickx component and I'd been limping along with that for a few months but I found it frustratingly unreliable. So I leapt at the opportunity to try your fork instead and I'm delighted to report that the heatmiser_ndc main branch has been working wonderfully for me. Thank you.

I don't have anything like as many stats as you (only 4x PRT-Ns into a UH-1, using a similar interface to your ATC_1000) so I've been able to reduce the scan_interval to 6s, with a COM_TIMEOUT of 0.25s and it's almost instantly responsive. Wonderful!

I tried installing your Additional-attributes branch recently but I had to roll-back rapidly when my stats stopped being polled and I couldn't figure out why. It was my live system so I couldn't spend much time on it right then. I stupidly forgot to make a proper note of logged errors. What I can say is that the integration was throwing out an error on start-up, on the lines of index out of range in line 348 of heatmiser.py (get_time).

My stats are simply numbered 1, 2, 3 & 4. Stat #1 would be unavailable but stats 2, 3, & 4 would have initial values and the additional attributes were present. As I mentioned earlier, no further polling was taking place.

Not very helpful, I know. I'll try again, when I have time, and feedback more useful detail but perhaps the scant info above will mean something to you in the meantime.

Thanks again for what is a very useful HA integration to me.

[NigelCoxon]

Hi Glad to know you find the code useful. I'm away at the moment, but will have a look in the next few days. I think I have seen this bug and fixed it on my system. I probably forgot to update GIT

On Fri, 28 Jan 2022, 18:52 washboy, @.***> wrote:

This is not a proper issue report (although I have been experienced difficulty with the Additional-attributes branch). It's a thank-you to @NigelCoxon https://github.com/NigelCoxon.

I saw your invitation to @tomtokic https://github.com/tomtokic, last November, to try your custom component. Tom had very kindly helped me get up and running with bretttickx component and I'd been limping along with that for a few months but I found it frustratingly unreliable. So I leapt at the opportunity to try your fork instead and I'm delighted to report that the heatmiser_ndc main branch has been working wonderfully for me. Thank you.

I don't have anything like as many stats as you (only 4x PRT-Ns into a UH-1, using a similar interface to your ATC_1000) so I've been able to reduce the scan_interval to 6s, with a COM_TIMEOUT of 0.25s and it's almost instantly responsive. Wonderful!

I tried installing your Additional-attributes branch recently but I had to roll-back rapidly when my stats stopped being polled and I couldn't figure out why? It was my live system so I couldn't spend much time on it right then. I stupidly forgot to make a proper note of logged errors. What I can say is that the integration was throwing out an error on start-up, on the lines of index out of range in line 348 of heatmiser.py (get_time).

My stats are simply numbered 1, 2, 3 & 4. Stat #1 https://github.com/NigelCoxon/Hass-heatmiser/issues/1 would be unavailable but stats 2, 3, & 4 would have initial values and the additional attributes were present. As I mentioned earlier, no further polling was taking place.

Not very helpful, I know. I'll try again, when I have time, and feedback more useful detail but perhaps the scant info above will mean something to you in the meantime.

Thanks again for what is a very useful HA integration to me.

— Reply to this email directly, view it on GitHub https://github.com/NigelCoxon/Hass-heatmiser/issues/8, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACVESZMJNA6C3BLKDT7C4HTUYLJVFANCNFSM5NBL6WCQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you were mentioned.Message ID: @.***>

[NigelCoxon]

I had already fixed the bug, but not updateed Git. The "additional attributes" branch is now updated. Please let me know if this works OK for you, and Ill merge into Main

[washboy]

Thanks for a rapid response. I've installed the update and it appears to be working fine.

I'll report back in a week or so (earlier, if necessary) to confirm that all's still good.

[tomtokic]

Hi guys,

I have just updated my system to use the "additional attributes" branch also. It's summer here at the moment so havent looked at this much of late. I am logging all errors, and see a few CRC errors from my thermostats. Not sure if that's a comms issue or code, but will let you know if I work it out. Not too much of an issue at the moment anyway. Cheers Tom.

[NigelCoxon]

Hi @tomtokic I too get occasional errors, mainly "no data read". The code recovers from this and seems to read the data OK on the next poll cycle. I have seen CRC errors too in the past. I have a Heatmiser Touchpad connected to my RS485, and I would think there must be occasional clashes between this and the ATC-1000, which will cause some errors. The Touchpad can display Comms Stats which shows about c1-3 lost packets per 1000 for each stat.

[tomtokic]

Thanks @NigelCoxon Good to know that it is not just my system. I don't have a touchpad and I suspect that it's more a matter the poll interval, but that's just a hunch. I am polling 8 thermostats every 60 seconds. I am going to push this out to 120 seconds and see if that improves things. I'll let you know if I dig anything up.

[washboy]

Hi guys

Just an update. Firstly, I don't know if I'm monitoring my log properly but I'm still not seeing any CRC errors, although I used to see lots when I tried bretttickx's fork. As I mentioned earlier, I have polling set to 6secs and a minimal 250ms com_timeout but then I do have only 4 stats. I do own a Touchpad btw, but it's not wired in. For me, things are functioning well so far. Not much help for you two, I know.

I need to do more testing but I'm not sure that the 7day times and temps are being reported correctly. I normally use 5/2 and that works well but, if I switch a stat to 7day and edit values at the stat, those changes remain as zeros in HA. Edits in 5/2 mode are correctly reported in HA.

I guess we're all working in Celsius but I tried switching one of my stats to Fahrenheit and some values went a bit screwy as a consequence. I realise that's using highly technical language and is somewhat over-detailed so I'll investigate further and report more usefully later ;-)

Reflecting on @tomtokic 's earlier request for a 'boost' function (seconded, btw!): it occurs to me that the addition of a service to write a value to the "Temperature Hold Minutes" (Hold Timer) in the DCB, combined with the existing climate.set_temperature service, would facilitate a simple HA script to set a boost temp plus duration. The target stat would automatically return to its programmed setpoint (or frost protect temperature) when the timer reaches zero.

I'd like to see a service to set a stat's time. I hate the Spring & Autumn chore of manually adjusting for DST and besides, the stats' time drifts badly anyway.