alexmohr / open-heat Goto Github PK

Hello,
I just got 15 BMP280 unstead of BME280...coming from China so I guess I will keep them. Could you please support BMP280 as well? Thank you

Have you ever had this issue? Get set temperature shows nothing. I had to restart HA and it happened.
Tried to restart MQTT, HA, nothing. Deleted and pasted.again the mqtt section in HA config, had it reappeared, restarted HA and disappeared again!

Originally posted by @Anycubic in #22 (comment)

either add remaining options to inital config or remove them and only keep admin user / password.

As soon as the mode changes on the webfrontend the new value should be pushed via mqtt.

Hello, I set up everything and all is working fine! Now I "only" need to add the valves in the loop, I have some Eqiva model without BT, could you please share a general connections/circuit diagram?
Thanks a lot for this project, it is looking very promising!

I don't know if you saw this behaviour before, valve was set to "off" in a proper way (checked also by sniffing the broker) but valve was actually not closed. Unplugged the ESP, replugged and I heard the motor spinning for quite a few seconds and actually closing the valve.
I'm not feeding the ESP with batteries, so low voltage is out of the question.

What about having longest deep sleep during night time (i.e. 2 hours) to increase batteries duration?

Add hint to reboot on webfrontend for settings which are only applied after reset

I'm testing a big radiator (1x2 meters), temperature tends to overshoot (i.e. set temperature was 21 and 22 reached + radiator was still pretty hot). I guess that maybe it's a matter of latency, which value can I try to play with (in RadiatorValve.cpp, right?).

web ui should enable selection of pins where motor is located

Hi, I'm working on a custom case STL. Could you please publish again the STL file? There are just Freecad files in Schematics folder but unfortunately Freecad crashes on startup on my Mac.

I'm working with last old-regulator build only. I noticed that battery values are always pushed when I switch on the valve, after that they are not pushed to MQTT every modem wake up but very erratically, while GET measured temp is very steady. Did you noticed that?

Modem sleep should be configurable via mqtt. Setting via webinterface is not supported.

If connection does not succeed config portal also should be started

After hours running fine in heat mode, in one radiator measured temp was keeping increasing like if the valve was dead. Lowered temp by 2 degrees and the valve not reacting. Shut off command was issued but still no noise from the valve. Recycled the power and I heard the valve closing eventually.
Going to serial log ...

Hi, it seems to me that both BMP280 release and merged DEV have a problem with deepsleep - it will never happen after first boot. Else enabled log and I always get stuck here:

[18:12:09][INFO]  [0000003946] Running MQTT setup
[18:12:09][INFO]  [0000003957] Device startup and setup done
[18:12:09][INFO]  [0000003968] MQTT topic: naima/, topic len: 6
[18:12:09][INFO]  [0000003975] MQTT subscribed to topic: naima/debug/enable
[18:12:09][INFO]  [0000003983] Sleeping for 1 ms

setting debug == false doesn't solve anything

Good morning, I see in the diagram BME280 Vin pin is wired to GPIO13 (I guess in order not to always feed the sensor so to save the battery). BUT in the readme also open window led is wired to GPIO13 by default.
I also tried to connect BME280 Vin to GPIO13 (I didn't configure it for window led) but it seems to me the BME is not working in this way.
Could you help me?
Thanks a lot
Giuseppe

One hour ago the ESP-12F stopped posting, to my surprise it was in config mode! I don't know if it is related, but double reset does NOT make the ESP-12F in config mode, while more than once I saw in terminal that after a single reboot it was going in config mode. Did you fix NodeMCU and broke ESP-12F?

Good morning, I just compiled and flashed last night build but I can't see any log anymore coming from the serial interface and after the first configuration no connection to WIFI anymore.
Flashed 2 different Esps but at no avail. Is there something broken in the build? Or is there some leftover in the previous flash which is creating issues? I use ESP-Home flasher to actually flash the bin.
Thank you

If battery level gets below 10%, disable modem wake and try to hold set temperature until battery is loaded again. This gives the user time to change batteries without the danger of overheating or undercooling too soon.

Add hint to reboot on webfrontend for settings which are only applied after reset

How would be hard to implement that? Maybe using a digital input (closed/open) and writing on the proper MQTT channel could be feasible?

Regarding the 18650 shield. I don't know which version you are using, I got 10 pieces of the V8 version which has a push button instead of the on/off switch. This is a real problem because when in deep sleep the ESP-12F drain so low power that the shield doesn't detect it and just shut down... do you know if the there is a way to implement the good old "on/off" switch? Is it maybe the J4 in the pic? I was not able to find anything on the net.
I was also thinking about to feed directly the AMS1117 pin with Vbat like in the pic, so bypassing the shield completely or at least partially.

It could happen that the furnace is switched off and the valve will try to open and open but of course temperature will never increase. This could also lead to servo motor damage. What about putting a check i.e. if in 30 minutes from "heat" status temperature didn't raise by 0.3 celsius then switch off the valve?

Configuration portal returns garbage binary data in some scenarios.

• Loading without browser cache works sometimes
• curl works

Fix configuration portal or replace library.

In general having a temp sensor close to the very same radiator is not a good idea...but I have several high sized radiators where the valve is at approx 2,5 meters from the floor level.... so there is a big delta, even 2 Celsius and more, between temps at floor level and where the valves are in those high radiators.
I was thinking that a possible solution could be:

1. moving the BMP/BME away and in lower position from the valve. Don't know if long wires could be a problem, also aesthetically not very nice solution
2. having a separate ESP/BMP sensor only fixed in a nicer position and pushing measured temps only (so the valve will take them via MQTT instead that from internal sensor). Adding a display and a potentiometer this could also became a real "room" thermostat (though I guess you need wire tap, battery will not be enough because of display...)
3. very quick & dirt, having the possibility to add a correction value (via web interface) to the measured temps so that it can be possible to deal with these problematic radiators like if they are normal position (i.e. I don't need to remember I have to set 2 C more than usual). In general this could only fix "problematic" sensors, i.e. I have a BME reading 1,5 C more than real temp.

What do you think? Thank you