controlling a mains-powered oven

Just make sure that you supply complete cycles, not an odd number of half-cycles, which will introduce a significant DC component.

--
Alan Adams, from Northamptonshire 
alan@adamshome.org.uk 
http://www.nckc.org.uk/

Folderol wrote on 8 Sep:

I was imagining software that would cause the heating element to switch on or off only when the sensor gets to, say, 10% above or below the desired temperature. Am I being electrically simplistic?

--
Jim Nagel                        www.archivemag.co.uk 
>> "from" address is genuine but will change.  website has current one.

No, not at all. For a domestic oven that's quite good enough. It only (potentially) becomes a problem with big industrial heaters where you want precise temperature control and the heating system itself has a minutes long thermal inertial, because when you reach your setpoint and switch off, the element will still be pumping out heat for some time, and vice-versa for cooling.

--
W J G

10% is a bit large, but no. This is how electric oven controllers have worked for a hundred years, give or take. Some folks would call this a bang-bang controller, all on or all off. Hysteresis prevents rapid oscillations.

The idea of PID control would be to supply power in an amount that equals the loss of the oven maintaining a more accurate temperature without the wavering of the bang-bang controller. In reality there is little need for it in an oven and they are not so easy to make without oscillations.

--

Rick

In comp.sys.raspberry-pi message , Mon,

7 Sep 2015 23:49:07, The Natural Philosopher posted:

You have not, as far as I know, met me.

--

   Web   - FAQish topics, acronyms, & links.

yes. but its more than good enough.

--
New Socialism consists essentially in being seen to have your heart in  
the right place whilst your head is in the clouds and your hand is in  
someone else's pocket.

One thing I would do is implement a hard thermal cutout: so a bug in your software doesn't cause your oven to emulate a blast furnace and set your kitchen on fire.

Theo

The cheap (domestic) oven I have already has this, so I imagine most domestics also have one. However I couldn't see one on the commercial ovens I have...

Gordon

I should rephrase that... a /resettable/ cutout. Thermal fuses and such are all very well, but you don't want to have to buy a new element - I suspect some cutouts are use-once.

Theo

Just use an oven with a built-in thermostat. It should not be a modern electronic thing but the old style big knob on the front with crude indications of the temperature.

Set the thermostat to the max temp you want to get, and switch the power to the entire oven using a solid-state relay. No real problem when the control sticks, the thermostat will limit the temperature. Maybe the oven even has a timer (mechanical tick-tock thing that you wind up to some time and that ticks back). That will limit the amount of time that it can heat itself.

I don't know what your application is, but e.g. for solder reflow on printed circuit boards with SMD components, the use of a cheap "kitchen oven" with an external controller is a wellknown trick.

My application (and that of the OP IIRC) is replacing the capilliary tube mechanical thermostat in a domestic oven with a Pi...

As well as the geekery I do I also run a small catering company/bakery and I'm looking to eventually have a nice overall control system for all my ovens - 2 commercial ones and a domestic one as well as the prover/retarder and possibly timers for my mixers too.

I do know of several people who've done the "toaster oven" into SMD reflow mods though - helped several with interfacing to a Pi and the software, etc. (I write & maintain the wiringPi library FWIW which interfaces nicely with a BASIC intepreter which I also wrote...) so for me it's now "eating my own dogfood", or rather bread in this case :-)

Using the existing thermostat in my domestic oven isn't good - as the replacement doesn't go as high as I want it to - the original would get the oven up to 250C - the replacement barely reaches 230C so that's one reason I want a new controller for it - might as well be geeky about it even though the controller will end up costing more than the oven :-)

Gordon

It's always more fun doing your own thing isn't it? :)

As an aside, I think you want a J type thermocouple. It should be adequate for the range (600C?), is fairly linear and you can get standard 'cold junction' offset amplifier modules.

--
W J G

Indeed..

Maybe, but it's too easy to just plug some easy to use stuff together and I know that a K type and a MAX31855 chip does everything with an SPI interface and I know these work as I've helped others get them going:

I'm also under the impression that J types aren't good in some high humidity conditions and another thing I am considering is plumbing the oven into the water supply to make steam...

(One of my commercial ovens already has this facility although it's fairly crude - washing machine type solenoid spraying a jet of water onto the plate in-front of the fans!)

Cheers,

Gordon

For Jim (OP's) benefit, this is the kind of controller intended for the 'toaster oven' type modification:

(though this particular project is a bit neglected - there are others based on the software)

I see no reason why this isn't applicable to a 'proper' oven. While the form factor might not be what you initially thought of, one option is to have a pre-existing device like this to do the control, and use a Pi as the user interface. That way you don't have to worry about 'disc full' burning your dinner (or your house).

Going along the lines of Pi-as-UI train of thought, OSPID is based commercial PID controllers like this one (and those substantially more expensive):

I don't know the field well, but I imagine if one of those has a serial port it would enable you to set the temperature profile without having to use the buttons and display on the unit as-is.

Don't forget to check that you can switch the current though - a cooker is on a 30A circuit, so you may need to beef up the switching.

Theo (for whom the 'toaster SMD reflow oven' is one of those projects on the eternal backburner)

They are also ten cents.

--

Rick

In article ,

Cookers yes, but most domestic ovens are 13A or less. (there are a few exceptions though - the ones used on Bake-Off are 4.5Kw ovens, but they cost over a grand )-:

My domestic oven has a 13A plug and takes 10 of those 13A. The 2 commercial ovens I have are both 13A (3.1Kw) devices too. One that I may Pi control after the first oven has 2 x 1.5Kw elements and I'd switch them separately (2 SSRs) - more to negate needing heatsinks on the SSRs than anything else.

This is an issue for the home based microbakery - most commercial ovens are three phase...

And if you think this is OTT... :-) Some commercial ovens now have USB sockets to let you connect up a PC to allow uploading of bake/temperature/steam profiles...

Gordon

Gordon Henderson wrote on 6 Sep:

Sorry to quote so much, but revisiting a thread from September.

Here, the project went onto the proverbial back burner for a while, but it's heating up again. I wonder if Gordon or others have gone further with it in the meantime and might have further wisdom to impart.

--
Jim Nagel 
   Somerset

Ah, I'd almost forgotten about this too - however the oven is Pi controlled and working really well. The PID still needs some fine tuning, but for now it's good.

Here is a graph of the oven under the existing capillary tube thermostat:

The three lines are 3 thermocouples - 2 are at the left back and right back, the 3rd (blue trace) is at the front. (All at the top, clipped to the grill element which is not functional) That was with the oven set to 180°C. The first dip is when I opened the door to stick a cake in, the others are when I checked it.

This:

is with the Pi controller - initial setpoint was 250°C - the dip is when the loaves were loaded in and note that the feeble little 2Kw heater couldn't get it back up to 250 before I turned it down to 210°C.

This:

is the same data once I'd learned how to drive gnuplot, zoomed in. Red line is the setpoint and the blue & purple lines are +/- 1°C (as are the cyan and yellow lines at the 2nd temp.)

This:

is a run this morning with 4 large wholemeal loaves - so the oven was effectively loaded with just under 4Kg of very wet dough and asked to get on with it... The blue line is the power to the heater (appears on the graph as either 200 < setpoint or 100 < setpoint when on), green is the actual temperature. Talk about being woefully inadequate to bring the oven temp back up to 250°C.... However it gets up to 210°C OK. The extra dip at the end of the 250 section is when I opened the door to move the tins about, and the next big dip is when I de-tinned the loaves and put them back in for a bit more time. The oven was then used at the same 210°C to cook a batch of buns...

I've done a few more tweaks to the PID and it's marginally better, but seeing that it maintains inside 1°C (with minimal over/under shoot) for the most part is good enough for me. (For now - I will tune it better when I get more time)

This is the controller:

The thicker grey wire goes to the SSR which is on a heatsink mounted inside the oven cabinet. It's wired in parallel to the existing thermostat so for now I can still use that if I need to. The 3 other wires go to the thermocouples.

I guess the real question is: Does it cook any better (or worse!) and I'm fairly sure that it does. Certainly my cakes appear better and more evenly baked - I still have to turn them, but it's a really cheap oven and without a lot of mechanical shenanigans to fix the airflow that's not going to change but I think having much short bursts of heat (once its at temperature) rather than longer periods of heating/cooling is much better - I see less scorching in the places that would normally get a bit too hot.

Cheers,

Gordon

Thanks for the update. It looks like a very successful project, and you can be proud of having done it yourself :)

--
W J G