Putting it all together and using it in production!

[jazzmonger]

I finally installed it in my stove a couple days ago.

Funny how things on the bench don't always translate to the real world. Honestly,
it was a disaster at first, I'd overlooked a couple key items. But persistence pays off. My wife thinks I'm nuts sitting and staring at the burn pot of the stove watching it light, go out, relight, etc.

I added a fire stoking routine that stokes the fire when the exhaust temp gets too low. I've found the sweet spot of Auger ON and off times as well as stoking params and can now keep the fire burning literally all day without it going out.

So, this is a huge win!

more soon.

That's 5 hours of blissful pellet burning and minimal heat output. Exactly what I was trying to achieve with all this effort. It's 40 degrees outside and the house is now staying a comfortable 68 degrees. Before in ECO2 mode on P4 power level after 5 hours the house would be 90 degrees and I'd be out of pellets !

[jazzmonger]

After experimenting fir a couple days,I added a stoking feature. The fire wants to go out if it goes below about 170 C at the exhaust. So I set up a sensor that catches that and "stokes" the fire. The stove has been burning
for over 12 hours now. I've been adjusting the burn params to get as stable a burn as possible at as low a temp as possible.

you can see me adjusting the burn params in this trace as the avg pipe temp falls from 210 C to around 170 C.
This is about as low as I've been able to get it.

here's a closeup of the stoking flag working.

Controlling a fire in a pellet burn pot is a delicate operation. Get the params off just a little too much and the fire goes out!

The TuyaMCU dp108 (pipe temp) only updates once every 60s. If I can get that sensor to update every 10s the feedback loop on the burn routines would work much better. There are only 2 ways to do this.

1. add a new thermocoupled exhaust pipe sensor. I have a few on order. But I may be out of esp pins unless I can repurpose a couple. These sensors take 3 pins. I could always add an esp32, but keeping everything in the onboard Tuya chip has a couple advantages. The biggest is that it is a commercially produced chip, so it will have the best reliability. Who knows what slave camp in China the D1 Mini I'm using was produced by. It could be just fine... or... not.

2. tap the TX line of the control unit and decode the data stream. That only takes 1 pin and I have that. This is the most likely, but decoding that data has been challenging. I've been staring at the spreadsheet for days trying to figure out what the data means.

more on this here #30

[jazzmonger]

This is working exceptionally well. It's 24 F outside and the stove has been burning 12~ hours.
The temp is set for 64 and the stove has been hovering between 65-66 F