PDA

View Full Version : [Question(s)] Best platform for oven controller



broadband155
04-22-2009, 06:05 AM
Hi All,

I really want to get my hands dirty with processors and programming and may have the ideal candidate. My family run a bakery. It has a fairly old electric oven which has really primitive and troublesome controls. The oven works as follows;

There are five 'decks', each of which has three heating elements for each of top and bottom of the oven. There is a single thermostat in each deck. You set the desired temperature with the thermostat and use a control graduated from 1 - 10 to set the 'heat' of top and bottom of each deck. This controller is a potentiometer which changes a timer to pulse power to the three elements. At 10, it is almost constantly on, whereas at 1, it is only pulsed on very seldom.

The control system seems primitive but seems to work except that the module which controls heat breaks fairly regularly and there are ten of them in the oven. It uses a triac to switch the load from the timer circuit and the triac fails.

The accepted solution is to replace the timer with a rheostat and the triac with a contactor. This does improve the situation somewhat but to me is still primitive and it strikes me that no two rheostats are the same, so you end up with different ovens behaving differently.

I want to build a control system that will take the temperature for each oven from at least one place in each oven (possibly more locations in future) and have two timers that are configurable to pulse two contactors for each deck. It should have an LCD readout of current temperature and heat settings and use buttons instead of pots for setting time. If one board can do all ten timers for five decks, all the better as it will integrate better.

Over time, I would like to do lots. There are several idiosyncracies with this oven. For example, the baker says that the ovens all work better if started from cold but we only have that luxury on a Sunday evening as they are used every other day for most of the day and night. It may be possible to find what the difference is and program for that as the controller will know the day and use the 'Sunday program' or 'weekday program'.

It would also be great to have the baker just set a program for bread, or cakes rather than use his own settings, where each baker will be different.

So, after all that, what I would like as a novice with microcontrollers and programming but years of experience as an electrician and with electronics is what platform to use and what programming language.

Thanks, Neil.

ScuD
04-22-2009, 06:15 AM
I guess an Arduino would be a viable candidate, though I have absolutely no experience with it myself whatsoever.

A pic could handle this easily, but since you're starting from scratch an arduino would probably yield faster results.

I would still go for triacs instead of contactors though, or maybe even better, SSR's. Granted, for temperature control you don't need high frequency switching, but you're still turning the things on and off at a fairly frequent rate, and even if it doesn't bother the contactor, it might bother the people working around it with it's constant clicks :)

Sienna
04-22-2009, 09:13 AM
Is there any reason that you wouldn't want to build 5 smaller controllers (one for each deck)?

If you limit each controller to a deck, that means that the common micros on the market will be better suited I think, as you are demanding less of them (input / output wise as well as processor wise) and giving yourself more expansion capability.

Plus, I think it would be kind of cool to have five small LCDs and control pads, one next to each deck :P

And I agree with ScuD, triacs or other solid state switching devices are better then a relay, as you are likely to burn out a relay in time.

jes1510
04-22-2009, 09:18 AM
I guess an Arduino would be a viable candidate, though I have absolutely no experience with it myself whatsoever.

A pic could handle this easily, but since you're starting from scratch an arduino would probably yield faster results.

I would still go for triacs instead of contactors though, or maybe even better, SSR's. Granted, for temperature control you don't need high frequency switching, but you're still turning the things on and off at a fairly frequent rate, and even if it doesn't bother the contactor, it might bother the people working around it with it's constant clicks :)


I'll second Scud's suggestions. The Arduino with something other than a contactor is the way to go. I believe you would see much better long term reliability by avoiding the contactor.

You may also want to consider an LCD for the Arduino for status information and it would let you do a nifty menu for setup and control. Something like this would be pretty nifty (Note the Arduino is not included)

http://www.ekitszone.com/index.php?main_page=product_info&products_id=2

broadband155
04-22-2009, 09:51 AM
Thanks for that. I think I will stick with contactors for now. The bakery is very noisy and one of the ovens has contactors as a trial and they aren't too bad really.

I had a look in the shop on the site here and there are a few different Arduino products but none of them look like what I need. I'm sure this is just my lack of knowledge. Is there a board with an Arduino in it that has plenty of I/O built in? Perhaps a developer kit? Also, is there an LCD module that I could use with it?

Thanks, Neil.

jes1510
04-22-2009, 09:58 AM
You would need an arduino, some kind of driver board for the contactor that you will probably have to build, and something like the LCD shield I linked above.

broadband155
04-22-2009, 10:04 AM
Great to have so many replies so quickly. I replied above before reading the last two:)

I have no real problem with five units instead of one except for cost and convenience but neither is too big a deal. One thing I did think is that there may be some interaction between ovens in the future. It is a long shot but there is heat transfer between ovens and for example, if oven three in being used on its own, in theory it will be harder to heat than if two and four are on also.

If this really is a noticeable problem, it would make sense to have the controller compensate for it.

On the contactors, you guys have me thinking now. It seems that a contactor lasts a long time in this application. Apparently several years. And they are cheap to buy. I guess I will look into cost and ease of installation and decide then.

One other consideration is replacement cost. Any electrician can easily replace a contactor.

Thanks again for all the replies.

Adrenalynn
04-22-2009, 12:08 PM
And they may fail closed rather than fail open. That could be a bad sorta thing in that environment - no?

ScuD
04-22-2009, 01:34 PM
Since I have no experience with industrial bakery ovens, what kind of power are we talking about here? 5kW? 500W?

Adrenalynn
04-22-2009, 02:21 PM
It's gotta be up there. A typical home oven could typically run 7200 watts peak. I've never even seen an electric industrial kitchen, so I would be interested to know.

broadband155
04-22-2009, 04:00 PM
There are three elements top and three bottom for each oven. They are around 1kW each element, so the contactor or other switching device needs to switch 3kW for each half of the oven.

I forgot to mention also that we have a three phase supply and the oven is wired for it, so it makes sense to feed a phase to each element. Otherwise, I would be switching 3kW on and off on a single phase very often which isn't very desirable.

Again, thanks for all the interest in my question and hopefully my first microprocessor project.

broadband155
04-22-2009, 05:21 PM
I was just reading a bit of this months Elektor magazine and it has 'Experimenting with the MSP430' and a companion article about 'beginning C programming'. It seems that C is a good language to learn but I don't know if it is beyond me at the start.

Can the Arduino be programmed with C, or would the MSP430 be a better option perhaps? I know there are so many different options out there that I could look forever but I didn't want to get into the wrong platform or something I get stuck on as too high a learning curve.

lnxfergy
04-22-2009, 05:45 PM
Just about any micro out there can be programmed in C. I'm not sure there is a GCC (free c compiler) for the MSP family. Even if there was, I'd recommend the Arduino/AVR architecture simply due to size of the user (hobbiest) community being so much larger.

-Fergs

Testing2000
04-23-2009, 10:09 PM
Greetings Broadband,

Something I know a bit about is thermal controls. I build, teach and maintain thermal control systems for testing space flight components. I have 15 thermal vacuum chambers and 9 thermal cycle chambers (atmospheric pressure) You don't need desktop control or programming to do this. The temperature should be in the baker's hands, Right?
I would be interested to know why they get better results from cold start though.
Not an endorsement but I use Omron controllers exclusively. There are half a dozen that work as well but wanted everything the same for operator comfort and reduced spare parts. All parts available from Newark. Bottom end is the Omron E5GN, about $220 each, LED readout of setpoint and process temp. Will work with 17 types of sensor. This feeds signal to Carlo Gavazzi SSR available for 220v 50A single phase about $90. Three phase 600v 25A about $220. You should have an ultimate over temp failsafe for fire protection but could use your existing control circuit for that as it will not be cycling at all under normal operation. These components will give you +- 2F temperature control with years of reliability and compensation between platforms hot or cold.

Adrenalynn, I had a Whal 4-20mA control SSR fail shorted about 10 years ago on a legacy (not mine) control system. Failsafe was in the control side not load. Fried a $40K unit.

Adrenalynn
04-23-2009, 10:19 PM
Wow. That's bad design. I hope someone got kneecapped over that. ;)

broadband155
04-24-2009, 08:14 AM
Hi Testing,

I hadn't thought of looking for something modular. I may be missing something but it is important in our application to be able to control the rate of heating, not just temperature. I had a quick look at the basic Omron one you talked about and it can't do that as far as I can see.

The bakers set the current units to anywhere between 6 and 10 which corresponds to the element being on perhaps 50% of the time or more than 90% of the time, depending on what is being baked.

In any case, this is a brilliant first project for me. I need to work on something practical where I can see real results at the end of it.

Thanks again for the feedback.

gdubb2
04-25-2009, 02:37 PM
Hey Broadband,
I would second what Testing said about the Omron stuff. It is very good and reliable. Looks like what you need is something with a ramp function, or ramp/soak. Omega Engineering has some controllers that do this.
http://www.omega.com/toc_asp/sectionSC.asp?section=P&book=temperature

I've used some of their stuff, and it's pretty good in my opinion. (30 yrs. as an Instrumentation Tech)

Unless you really feel the need to feed your hunger for knowledge, why reinvent the wheel for a commercial app.?

Good Luck
Gary

Testing2000
04-26-2009, 02:04 AM
I guess I don't understand commertial baking. In making bread I have always preset the oven. What are your bakers using for feedback

broadband155
04-26-2009, 08:48 AM
Testing,

I agree. I had always thought that an oven got to a temperature and maintained it. I couldn't and still can't really understand fully why you need a rate of heating.

Just to be clear though, they do set the ovens to a temperature, say 220 degrees celcius and wait for it to heat to that temperature before putting food in the oven. I checked with one of them only yesterday to make sure he actually uses different heat settings and he says he does. Bread is usually at nine and pies for example at seven.

My logic (which is obviously flawed) is that if you decrease the rate of heat, it won't maintain the temperature but it does I think. I do know that if you decrease the heat with bread in the oven, it takes far longer to bake it and it won't turn out very good.

Perhaps it is partly to do with the thermostat used. The 'probe' is about 300mm long and its a fairly big chunk of metal. It has a capillary from it to the gauge and limit switch. I haven't tested it extensively but i'm sure it doesn't react very quickly to the temperature in the oven, so if the controls are set to nine and the stat is calling for heat, maybe the oven heats quite a bit hotter than the set temperature before the stat catches up and cuts it out. If it is set to six or seven, it isn't as quick and maintains the temperature more evenly.

Then again, maybe that is all rubbish and there is another explanation. I would say though that when a deck is full of bread, there are 18 loaves in it and that is a lot of material to heat. It must absorb quite a large amount of heat in total.