Loving my BrewPi so far!

[Matt in Birdham]

Is BrewPI still only aimed for fridges?, so it controls the chamber temperature. I cool/heat the FV directly, so only have one temperature input, the beer. Anyone using BrewPI for direct control ? The way i see it, as there's sigficantly less lag time when you are cooling/heating the beer directly, a simple STC1000 type on/off thermostat rather than PID is sufficient. But like others, i have a spare Arduino and kit to have a play with this.

It doesn't know where it is or how it is heating or cooling. It has some PID type algorithm that learns rates of change and various other parameters so should work fine for your direct method.

I'm not sure this is entirely correct. The BrewPi controls beer temperature by targeting a fridge temperature, so it needs those two separate temperature inputs. I believe it runs three PID algorithms at once; fridge cooling, fridge heating and "beer to fridge", which calculates where the fridge temperature needs to be to get the correct beer temp. It's quite a clever system, that solves the issue of overshoot when controlling a fridge directly and I think it is quite specific to that setup. Having said that, you could probably use the "fridge constant" mode and use the beer probe as "fridge" temp, which would then just use the heat/cool PIDs, which could be an improvement over the STC1000.

[T5FAU]

Is BrewPI still only aimed for fridges?, so it controls the chamber temperature. I cool/heat the FV directly, so only have one temperature input, the beer. Anyone using BrewPI for direct control ? The way i see it, as there's sigficantly less lag time when you are cooling/heating the beer directly, a simple STC1000 type on/off thermostat rather than PID is sufficient. But like others, i have a spare Arduino and kit to have a play with this.

It doesn't know where it is or how it is heating or cooling. It has some PID type algorithm that learns rates of change and various other parameters so should work fine for your direct method.

I'm not sure this is entirely correct. The BrewPi controls beer temperature by targeting a fridge temperature, so it needs those two separate temperature inputs. I believe it runs three PID algorithms at once; fridge cooling, fridge heating and "beer to fridge", which calculates where the fridge temperature needs to be to get the correct beer temp. It's quite a clever system, that solves the issue of overshoot when controlling a fridge directly and I think it is quite specific to that setup. Having said that, you could probably use the "fridge constant" mode and use the beer probe as "fridge" temp, which would then just use the heat/cool PIDs, which could be an improvement over the STC1000.

Thanks, it was the mode option and probe settings that i was after, ill give these a go once i have it setup.

[aamcle]

If you want to mash and boil why not try CraftbeerPi http://www.craftbeerpi.com

Aamcle

[Matt in Birdham]

Thanks, it was the mode option and probe settings that i was after, ill give these a go once i have it setup.

Well just to be clear I don't know how well/if that will work, I was just thinking aloud really. It makes sense to me though

[AnthonyUK]

I'm not sure this is entirely correct. The BrewPi controls beer temperature by targeting a fridge temperature, so it needs those two separate temperature inputs. I believe it runs three PID algorithms at once; fridge cooling, fridge heating and "beer to fridge", which calculates where the fridge temperature needs to be to get the correct beer temp. It's quite a clever system, that solves the issue of overshoot when controlling a fridge directly and I think it is quite specific to that setup. Having said that, you could probably use the "fridge constant" mode and use the beer probe as "fridge" temp, which would then just use the heat/cool PIDs, which could be an improvement over the STC1000.

I think you may be able to use one probe and the beer constant mode.
If the single probe isn't supported I think you can set Kp of the fridge probe to a value where it is basically ignored.
I haven't tried it though so this is just a guess.

[Matt in Birdham]

I think you may be able to use one probe and the beer constant mode.
If the single probe isn't supported I think you can set Kp of the fridge probe to a value where it is basically ignored.
I haven't tried it though so this is just a guess.

I think you'd want to go "fridge constant", because that temp is directly attacked by the PID. Beer temp is set by varying fridge temp, so 2 probes would be required. You would of course need to have your beer probe configured as "fridge" for it to work.

[Andy]

If you want to mash and boil why not try CraftbeerPi http://www.craftbeerpi.com

Aamcle

Cheers for the link, will check that out.

[Matt in Birdham]

In the interests of full disclosure, I have to report that my BrewPi cr*pped itself yesterday!
I have no idea exactly what happened, but I noticed when checking on my brew in the evening that the graph only extended up to lunch time yesterday (about 6 hours missing), although the web page was still responsive. Upon inspection of the unit, it was clear that temperature was no longer being maintained and my beer had got to 23C (not a huge drama, since I was in the middle of a ramp from 19 to 22, having achieved my FG 2 days ago - could have been a lot worse early in the fermentation though!).

Reboot didn't help and I could no longer connect to the RPi via http or ssh, so I brought it inside and connected to a monitor - just page after page of trying to fix corrupted files on bootup, which I killed after about 20 minutes. I decided to reinstall the whole lot, which was a right PIA at midnight last night, but I got it done in about 1.5hrs, then set on a beer constant of 22C and it seems to be working well again (all on a new SD card, obviously).

It would be nice to understand what went wrong here, since I haven't touched the unit at all. It seems like something went FUBAR on the RPi side, probably the SD card.

The lesson here is that if I had imaged a fully installed and working RPi SD card, I could have resolved this situation in about 5 minutes and not 2 hours at midnight on a Sunday, so a couple of spare SDs on order now and I will snapshot my latest one. I do hope though that the RPi isn't going to make a habit of this as the outcome could have been much worse (I don't think this beer will suffer at all, fingers crossed).

Also worth noting that the spark can run without the Pi, so I could have just left that powered. This does not explain why the temp control seemed to fail though - but the spark logs did show a lot of failed writes to the RPi serial port so maybe this affected it in some way? At least when I reinstalled the Pi side of things, all of the photon settings remained intact (devices, control algorithm tweaks etc).

[Aleman]

Yeah, its a very good tip to Image the card once you have the PI working the way you want to.

I've had the PI trash itself twice through power failures causing corrupted SD cards. I'm still thinking of getting a surge protector/small UPS to guard against that. . . .but the image works just as well.

The Arduino usually proves to be rock solid compared to the Pi, and happily has kept the temperature where I wanted it to be on the failure of the Pi.

Windisk Imager is pretty easy to use to image the SD Card.