My homemade digital temperature controller!

[robgunby]

OK, it's becoming clearer. Forgive my ignorant questioning, but, if I connected, for example, one of these http://www.amazon.co.uk/AGPtek%C2%AE-K- ... controller to a fan and some ducting, would it do the same job? If not, what additional functionality does the arduino provide please?

Just trying to wrap my head around all the options - your patience with my technical ignorance is appreciated!

[aris]

If you do a search here, Steve hayes did exactly that.

viewtopic.php?f=43&t=1690

[robgunby]

Yeah I had a recollection Steve had made a simple one. Does the arduino offer any extra functionality? You mentioned the PID on its own was difficult to train to the smoker, I guess with the arduino do you have more control over the learning mode?

[aris]

Arduino can do anything you want if you know how to program it.

[reinhart36]

I've gotcha (I think - I tend to leave such matters to the experts!). So, the arduino isn't strictly necessary for a PID controlled setup, but provides an easy to use interface for the control of said PID unit?

Yeah, Aris explains it well. In general, digital devices that control everything in our lives all have a core "computer" chip at the center of them. If you want to build a computer, anyone can buy a chip and build a board around it that gives it power and a means of attaching to sensors, displays, etc.

Arduino, Basic Stamp, Rasberry Pi, and some other brands are basically an already-built board with a computer in it. They still need to be programmed, but are much simpler to use than building from scratch. You should definitely try it out - it's how I got my start.

[reinhart36]

Incidentally, for my project, I used a Microchip PIC18F controller chip and just built a board for it to go into. The chip retails for about 1 pound if you're buying only one of them. Then, of course, a few more are needed for the rest of the parts. Electronic components are amazingly inexpensive!

[robgunby]

Thanks guys, most helpful.

Engineer matey is on about building me a PID from scratch which will save me a chunk of cash - way beyond my expertise so I'll leave that to him!

I think my first prototype will have to be a PID only unit, as my programming knowledge is not up to scratch - I know some very basic python but not enough to consider starting a project like this yet.

[aris]

I've gotcha (I think - I tend to leave such matters to the experts!). So, the arduino isn't strictly necessary for a PID controlled setup, but provides an easy to use interface for the control of said PID unit?

Yeah, Aris explains it well. In general, digital devices that control everything in our lives all have a core "computer" chip at the center of them. If you want to build a computer, anyone can buy a chip and build a board around it that gives it power and a means of attaching to sensors, displays, etc.

Arduino, Basic Stamp, Rasberry Pi, and some other brands are basically an already-built board with a computer in it. They still need to be programmed, but are much simpler to use than building from scratch. You should definitely try it out - it's how I got my start.

What kind of algorithym did you use on your device to stabilise the temperature? PID? I'm finding it hard to tune this as the input and resulting outputs are so laggy.

[aris]

Thanks guys, most helpful.

Engineer matey is on about building me a PID from scratch which will save me a chunk of cash - way beyond my expertise so I'll leave that to him!

I think my first prototype will have to be a PID only unit, as my programming knowledge is not up to scratch - I know some very basic python but not enough to consider starting a project like this yet.

It should work. What i'm working on has a bit more to it. Right now it has:

- 4 temperature probes (3 food, 1 for the pit). I could easily add many more, but that's what i'm working with now.
- External temperature/humidty sensor - might use this data in the cook calculation - but interesting nonetheless.
- Relay switch for a fan (forced air)
- Bluetooth connectivity so it sends all the data to an Android Tablet every few seconds. Nothing fancy just yet - just raw data to the screen via bluetooth SPP.

Other ideas I had which are very easily implementable:

- Kick of a solenoid to shake the charcoal basket at intervals, and/or when the pit temperature drops and is not responding to the forced air
- Ir proximity sensor to monitor the lid so I can have the forced air sleep for a few minutes after removing the lid.
- LED countdown timer on the BBQ itself - estimate the ETA of ready food at a glance.
- Possible wifi connectivity - not that much different to bluetooth, but that is for later.

[reinhart36]

What kind of algorithym did you use on your device to stabilise the temperature? PID? I'm finding it hard to tune this as the input and resulting outputs are so laggy.

I started with PID, but evolved a lot from there in a variety of ways.

Are you familiar with Ziegler-Nichols tuning method? I suggest starting with only proportional control, and find out what your stability limit is for it. Then back it down by, say, 50% and introduce the integral gain at maybe 10% of the proportional gain.

Don't use a derivative term at all until you get lots and lots of runs made with just the other two. Derivative is "icing on the cake" but will totally hose you if it's in the mix when you're getting started.