I'm trying to make a film processor. The company Jobo make these kind of things but kinda expensive new ($3000-$4500).
Basically what they do is rotate a tank on some bearings and heat the bath with the element inside it (see picture).
So I'm trying to make this bath right now. I know I'll need a Temp Sensor, PID controller, SSR (Solid State Relay) and an element to heat up the bath. However I have a few questions:
Where can I get a heating element? And how does it all connect together.
How do I define the temperature of the bath with the Arduino?
Can I get the water up to temperature with a precision of +- 0.2°C precision. It should be very constant and not dropping or rising very much.
For the rotating part I'll use a stepper motor and controller. Been there before with CNC's so that shouldn't be a problem.
First you need to build the bath, then figure out how much power (in watts) is required to maintain the water temperature. After that the electronics will be straightforward.
I used to do a lot of film processing and 0.2 degrees C precision is definitely overkill. If you can keep the temperature constant to 1 or 2 degrees everything will be fine.
Just curious (even if this is slightly off-topic): I assume from the temp you are using and the level of precision that you need that you are doing color work. Also, the picture looks like a sheet-film processor. So my question is: do you have a freezer full of LF color film, or are you betting that manufacturers will continue to make it? I've been dreading the day when they stop producing the stuff and my beautiful 4x5 woodie turns into nothing but a film prop. =(
Thats a jobo processor and i used to use one.
As said 0.2 is overkill, but can make a difference for colour but using one of those processors i found about 2 to 3 deg drop at pour out time anyway.
For 0.2 deg you will need to use platinum resistance thermometer.
Heater is difficult, i used to use aquarium heaters but these days they are self contained and limited to 28 deg.
Caswell plating do or did a 200 W quartz submersible heater which does the job.
Failing that could try one of the little cup boiling elements sometimes found in camping shops.
Sheet processor, the magnetically coupled drum can be changed for a spiral wound one for film.
Edit
Caswell also do hollow floating plastic balls which are useful for reducing surface heat loss.
With a pid controller you can set the temperature directly, controllers are available for platinum and various thermocoulpe sensor.
Alternativley you could use the arduinos pid library and roll your own.
RobRothman:
Just curious (even if this is slightly off-topic): I assume from the temp you are using and the level of precision that you need that you are doing color work. Also, the picture looks like a sheet-film processor. So my question is: do you have a freezer full of LF color film, or are you betting that manufacturers will continue to make it? I've been dreading the day when they stop producing the stuff and my beautiful 4x5 woodie turns into nothing but a film prop. =(
Rob Rothman
Ha man, film is far from dead. I guess you are a little older than I am (only 21 here).
Kodak isn't really into film anymore but they still produce lots of it because there are people who need it. I know tons of wedding photographers that went back to film. And lots of cinematographers use Kodak Vision3 film. For movies like Fast Five, The Secret Life of Walter Mitty, The Grand Budapest Hotel, The Wolf of Wallstreet..
Ilford is steady as they go. Spokesman Simon told us that they won't be quitting the scene any time soon and that film probably will still available within 100 years.
Adox/Agfa are still making film and paper.
Film Ferrania starts producing AGAIN since they closed end this year (2014). Will be making color reversal film in all formats.
Cinestill is also a new guy, they make cinema still film from Kodak Vision3
Fuji is closing down it's film part but still lots available until the demand lasts
So no I don't own a freezer full of film. I'm shooting every week, 99% on film. Most of my work is for modelling agencies. You can check my work at www.jessestr.be
Don't buy expired film and get some new film.. It keeps the business alive.
I think I'm going to get an arduino, stepper driver and motor for that part. Use Jobo tanks with magnets to couple them to the motor.
And use a PID Temperature Controller from eBay where I can set the temp myself. It will be far more accurate. Only thing I can't find out is where to get the actual heating element that could lie down in water.
Check out aquarium heaters. They come in glass and titanium, i would pick titanium for this application (glass has more thermal mass). Stay away from electronic ones, just get the kind with a physical bimetallic strip thermostat. Set it as high as it goes, then control the power to the heater with an SSR (Solid State Relay). You can cycle the power on and off at 100ms intervals for very precise control.
tylernt:
Check out aquarium heaters. They come in glass and titanium, i would pick titanium for this application (glass has more thermal mass). Stay away from electronic ones, just get the kind with a physical bimetallic strip thermostat. Set it as high as it goes, then control the power to the heater with an SSR (Solid State Relay). You can cycle the power on and off at 100ms intervals for very precise control.
another idea is to use your coffee maker water heater and a continuous pump.
and last, and least, the dollar store. you should find a 99 cent plug in unit that will boil water in a cup. couple of those, one to switch on full to help get to temp, then one that is on the SSR to maintain temp. much better to have one correctly sized to maintain temp than try to use an incorrectly sized one and try to make up for the problems in software.
Dish-washer heating element perhaps? They go to 66C.
Does the element have to come into contact with chemicals BTW?
As for the PID you'll need to output a slow PWM signal (period 10 seconds or so?).
(The thermal mass means a slow rate will still give good control, and saves switching
the mains too rapidly - RFI)
Tuning the PID will take time, of course. Having a means to graph the temperature
against time automatically would be very handy for this purpose.
tylernt:
Check out aquarium heaters. They come in glass and titanium, i would pick titanium for this application (glass has more thermal mass). Stay away from electronic ones, just get the kind with a physical bimetallic strip thermostat. Set it as high as it goes, then control the power to the heater with an SSR (Solid State Relay). You can cycle the power on and off at 100ms intervals for very precise control.
Modern ones in uk cannot be set high enough these days, some iv seen have a fusible safety link.
tylernt:
Check out aquarium heaters. They come in glass and titanium, i would pick titanium for this application (glass has more thermal mass). Stay away from electronic ones, just get the kind with a physical bimetallic strip thermostat. Set it as high as it goes, then control the power to the heater with an SSR (Solid State Relay). You can cycle the power on and off at 100ms intervals for very precise control.
Modern ones in uk cannot be set high enough these days, some iv seen have a fusible safety link.
Try it and see i suppose.
Ask your goldfish for permission before trying this experiment!
Thanks for all the help and the links to the heating elements.
The chemicals will never touch the heating element. It's only to warm up a water-jacket where the chemical bottles will be sitting in.
So now I see every part coming together. Would it be easy enough to program it myself? Programming itself is not the issue, but understanding the PID and make it run efficient.
Been reading something about Point Value, Set Point and the duty cycle.
Should I get a ready to go PID Controller on eBay or just do it myself? Just need to be sure it's accurate enough. To let's say 1°C.. Can always go more accurate.
Ready made PID regulators often have autotuning. That can help a lot if you are new to regulation. They also often interface directly to SSRs and to common temperature sensors like thermocouples and pt100.
If you still want to bake your own, a period time of about 10s is nice. That will give you 1000 zero crossings (@50Hz)
