Generic thermostat behaviour, also smart_thermostat and Firtz Dect 200 Plug

Hi Folks,

I run a generic thermostat and I get this behavior:

grafik

and thats the code for it i run

- platform: generic_thermostat
  name: officeheating
  unique_id: 46124de8-b4b3-4867-9f11-6a6077f11445
  heater: switch.buero
  target_sensor: sensor.office_temperature
  min_temp: 10
  max_temp: 26
  ac_mode: false
  target_temp: 21.5
  cold_tolerance: 0.5
  hot_tolerance: 0
  min_cycle_duration:
    seconds: 6
#  initial_hvac_mode: "off"
  away_temp: 10
  comfort_temp: 26
  precision: 0.5

so what I would like is it starts heating a K above the adjusted temp and stopps heating below the adjusted tem, so the bouncing is not so bad, or just starts pulsing on off like 10 sec on 20 off 10 on and so on, so it kind of changes the duty cycle to just to just end with a nice constant room temperature.

what am im doing wrong with the generic thermostat? I dont think this is the intended behaviore?

Thanks
Manne

I think its the cold and hot_tolerance, i would need to set it to a negative value, i dont know if that works but I try it out right away.

Maybe try this:

1 Like

is it possible to set negative numbers for the hot and cold_tolerance?

I would like to reduce my temp bouncing, like the heater should start 1 degree before the target while coming dwon, and it should stop 1 degree early coming up…

thanks

climate:

  - platform: generic_thermostat
    unique_id: heat_p1_7c
    name: Heat P1 7C
    heater: switch.athom_p1
    target_sensor: sensor.mho_7c_temp
    min_temp: 15
    max_temp: 21
    ac_mode: false
    target_temp: 18.5
    cold_tolerance: 1
    hot_tolerance: 1
    min_cycle_duration:
      seconds: 5
    keep_alive:
      minutes: 2
    #initial_hvac_mode: "off"
    away_temp: 16
    precision: 0.1

18.5+1
18.5-1

while im here, i did run it a cold night at heating to 18°C

i changed the

    cold_tolerance: 0.5
    hot_tolerance: 0

I had in my code to

    cold_tolerance: -1
    hot_tolerance: -1

here is the result:

It sort of wokrs, but the coded logic gets a bit confused… :slight_smile:

left 4 cycles where with my old code, rest is with the new code.

and I think I have an issue with the Fritz Dect Plug that does the measuring and switchting. every time the heater is on for more than just a few seconds the temp drops by 2 K, every time the heater goes to off, the Temp rises by 2 K…

I may be will have to use two seperate units.

Anybody knowts a reliable temp sensor and a switch handling continus 2,5kw, both prefered to run with wifi, tasmota and powersupply from the grid (well for the switch thats obvious, but for the temp sensor too)

So I got the smart thermostat set up an running, its in its autotune phase, now, dont get exited if it does not start, you can monitor the satus in the dev tools conditions filter for the smart thermostat, it will say some considerabel amount of time that the autotune did not jet start. it took way past several hours on my HA to start the autotune also it was already way colder that it was set for

and

after some more consideration, i come to the conclusion that the current draw through the socket is not the reason for the 2K drop i see after about 5 minutes in the temp curve of my heater setup.

I suspect its a thermal thing. The heater starts to move the air in the office by thermal velocity, and by that cold air is flowing over the sensor and generates that drop in ambient temp in the curve, similar after the heater stopps, that airflow stops and the ambient egalizes and shows the true temp, I may have to move that socket to an other place with no cold air flow form upstaris right over the sensor.

i think i need to roll back my previous statement, the 2 K temp jumps do have something to do with the current drawn though the plug.

the smart_thermostat made that clear, it happend every time there was a significant change in current flowing through the plug.

every jump in the picture represents the smart_thermostat switching the heater

so i will have to “modify” the measurements, when the switch is active AND the current is above 10 amps or so, jet to be evaluated, the temps reported back have to be modified by +2 K

Hi there,

I run a heater by using a fritz dect 200 plug, it also provides the temp reading that does controll the heater.

It produced a 2K temp jump when the heater was running, I first suspected the current to be responsible, but then compared time between jump and switching. Then came to the conclusion it must be a thermal think like the radioator gets the air moving, and therefore cold air flows over the sensor, silly me, that would have triggerd a step by steb incremental change, not a straight away 2 K jump.

Then I changend from generic to smart thermosat, and it became bluntly obvious, at least obvously enought for me to go back to my original conclusion

every jump is a toggle action of the heater.

since it goes down it cant be heat inside the switch.

so, how do I fix it? because this way my smart thermostat is kind of getting silly by the jumping main control unit.