DIY water bath

Hi!
Slowly but surely, I'm building my controlled water bath for processing my color film (with a PID).

A friend of me who does home renovations professionally took a quick look at my electric plan. Told me that it seemed to be ok.

It would be much appreciated if someone who has built its own DIY controlled water bath could take a look at my electric plan. Cannot be to safe.

Thanks!

Note : I guess that the way I have drawn the wiring for the switch is somewhat misleading since the live wire will "come in" on one screw while the "outgoing" wires (relay + PID) will be plugged on the other screw.

I haven't built one (so take what I say with a large grain of salt) but unless your water bath is very large, isn't a 1500w element overkill? My concern is that it would cycle on and off so frequently that you'd experience too much variation. Just a thought; I'd be interested to hear from someone who actually knows.

Dear Draveur,

I won't tell you it is safe. I will tell you that it looks a lot like my setup but mine is wired such that the heater can't come on unless the pump is running.

Good luck,

Neal Wydra

Hi, I'm not an electrical specialist but I've worked on design of such things with folks who are. I have a few comments based on my experience in photofinishing.

First, you are using a proportional-style controller, but using it to turn on a relay. Now relays are essentially remote-control on/off switches, so they don't have things like half power, etc., they are either on or off. So I don't think a proportional control is the proper way to operate a relay - you need a controller that turns the relay on or off. A relay IS a good way to operate a large heater like your 1500 watt unit. However... I share mgb74's concern about too much temperature variation. I think, for tight temperature control of a relatively small water bath, you'd be better served by a control system that can vary power to the heating element. Now I don't know for sure, but I suspect that they may be costly in a size to handle your big heater.

A few words on controlling water temp: the system design, in general, is fairly critical. You want to have a circulation pattern that minimizes the temp variations - for example, if you have a "dead spot," the temperature there will tend towards ambient. Even given good circulation, placement of the sensor can be crucial. For example, if you need only a slight increase in temperature your system will turn on the (large) heater at full power - water right next to the heater will begin to heat right away. But if it takes, say 5 seconds for the warmed water to reach the temp sensor, you may have done enough heating to significantly overshoot your aim. It might seem like the solution to this is to put the sensor right next to the heater - the problem with this is that the sensor will see the heat right away, and might turn off the heater right away. So if you are starting with a cold tank, it might take a long time, with many on/off cycles, to warm up. My point is that there is more to a good design than meets the eye.

Regarding safety, I am by no means expert in this sort of thing. But I think a lot of this depends on the heating element itself, as well as having a good earth ground for it. Internally in your control box I don't see any fuses. So if you have an internal short circuit (even in the heating element), you are relying on the electrical supply system to blow a circuit breaker. I wouldn't want to count on this, especially since I don't have any idea of the amperage capacity of that circuit, and it may be more than your controller wiring can handle. So I would want a fuse built into the control box. And if it's a metal box, be sure the chassis is grounded (you would expect that the GFCI would sense any current leaks and break the circuit, but what if your controller is eventually moved to a non-GFCI circuit).

As a note on your diagram, the +/- notation is really for DC systems. You are (presumably) using AC, where the standard U.S. terminology for standard household current (~120 VAC) is to speak of a "hot" line and a "neutral" line, see here:
https://en.m.wikipedia.org/wiki/Electrical_wiring_in_North_America

Best wishes on the project. Hope this doesn't scare you off.

I get the point. Since the SSR is "just" a switch, it is not capable of delivering say 25% of power if the PID think that 25% will prevent overshoot.

I have read that PIDs work on cycles: it the PID think that it should give only 50% of power to prevent overshoot, than, in a 5 sec. cycle (for example), it will heat only during 2.5 seconds (at 100% since the SSR is only a switch). If this is true, than it would mean that the heating element would not heat long enought to boil the surrounding water.

So the "traditionnal" setting (like this one : http://vwblvd.com/?p=1563) would work with a standard SSR and (too big) heater. But I might be wrong.

Now, I have done some research. There seems to be "proportional control SSR" like this one. Also : Solid state variable relay --zero-crossing control. Are these a solution?


Good point as for the localization of the sensor. The best way will be to make tests with an electronic thermometer.

I entend to put all the wiring in a plastic box and the water bath itself will be of plastic or plexyglass (if not wood with a coat of resine for fiberglass). The heating element will be installed in an electrical box (PVC) to protect the connections.

The GFCI will not be "in the wall", but a unit connected on the main line. So (unless I'm wrong) I should not need a fuse. In other words, the GFCI is part of my "system".

A again, thanks for your input. It does not scare me off. I will only take more time to build my water bath.

R.

I've built 3 PID controlled installations: one to control my drying cabinet, replacing the old electro-mechanical control unit, one to control my development bath temperature, one for sous-vide vide cooking.
Your electrical scheme looks OK to me.
A PID is not a proportional controller: it swithes the SSR on and off for certain periods of time, depending on the difference between the set temperature and the temprature measured by the sensor. The PID "learns" from the way the on/off cycles change the measured temperature and becomes more stable and accurate over time as long as it is switched on and circumstances stay the same.
1500 Watt seems like an overkill to me: I found that PID's work better with low-power heating elements.
I prefer NOT having an on/off switch but instead unplugging the entire system when not in use.

Ron789 said:

I prefer NOT having an on/off switch but instead unplugging the entire system when not in use.

Click to expand...

With a 2-phase On/Off switch right after the power cable you would be second best concerning electrical safety, just behind pulling the mains cable.

Ron789 said:

. . . I prefer NOT having an on/off switch but instead unplugging the entire system when not in use.

Click to expand...

Switches are intended to turn loads on and off: power cords and outlets are not.

Ron789 said:

I've built 3 PID controlled installations: one to control my drying cabinet, replacing the old electro-mechanical control unit, one to control my development bath temperature, one for sous-vide vide cooking.
Your electrical scheme looks OK to me.
A PID is not a proportional controller: it swithes the SSR on and off for certain periods of time, depending on the difference between the set temperature and the temprature measured by the sensor. The PID "learns" from the way the on/off cycles change the measured temperature and becomes more stable and accurate over time as long as it is switched on and circumstances stay the same.

Click to expand...

The guy who's done it should know more about it than me, so you probably should listen to Ron789.

But... I wonder if this style of controller is a special case; I looked up PIDs in an electrical handbook where they showed some circuits based on op amps. I had (erroneously) presumed that the output always varied voltage, but apparently these units can vary frequency, which seems to be no problem for solid-state relays. I guess the rule is to just follow the manufacturer's instructions.

And as Ron789 says, they seem to also have a self-tuning function built in.

Why not just get a prebuilt one like a phototherm model 14 bath or even just the controller/heater/pump unit?

Perhaps I am missing something but this seems like overkill to me unless for higher volume professional processing. Unless using my ancient Jobo for some photography I use a water bath consisting of a large plastic basin and a 24" tropical fish heater when processing photo and movie film. My darkroom is in my basement, which is cool in the summer and cold in the winter and my simple system works well for me. However, I should note that my volume of E6 is so few that I no longer process it myself so I do not need extremely high temperatures.

Aquarium heaters can be tweaked to give higher temps than the factory settings. I did it once to try out a cute still design.

You can set the heater to about 5 degrees warmer than you need and control it with a light dimmer. That way, the on/off thermostat never comes into play meaning no problem with hysteresis.

I'd guess the cheapest and safest solution would be to have a container of water that was a bit too warm and wait for it to cool down. There must be a minimum volume of water that will stay within half a degree for 4 minutes.

One more note on aquarium heaters. My homage to a Jobo CPE uses an inline aquarium pump and inline 300W heater. To do C-41 I had to remove the temperature controller inside the heater to get over 99°F.

Neal Wydra

Hi.

There are many commercial solutions, of course. But I like to DIY once in a while. You learn stuff. As for tweaking aquarium heaters, I think it represents a real risk. Another reason why I'm want to build my own water bath, is that I have to ship to another city my C-41 films. The shipping fees are so high that home C-41 becomes a real alternative (it will repay the project within a year). And I might as well to do a "pro" job so it will last and it will make my film processing results (C-41 + B&W) more even.

I got rid of the switch: my plan got much simpler! Thanks for this one!

What would be the "best" heating element for optimizing the auto tune fonction of a PID?

I have found a Camco 2000W 240V heating element. Unless I'm wrong, that would give me ~500W. Is that still too much? Or even this one: 220V 750W (110V, 190W). Not enough?

How about "cup heaters"? This one is 220V 800W (110V, 200W). I could in this case, plug 2 of them in opposite position for better even T°C (+ water pump). But at a glance, I would say that they will be more difficult to install.

(from a quick search, I just cannot find a low watt 120V heating element. Any suggestions?)

Thanks to all. I really appreciate your comments. I have tried electronic forums where I got rather useless comments...






So the PID is the right way to go.

There's a lot of interest just now in cooking "sous vide" essentially poaching in long slow water baths with the food in plastic or whatever. There are a lot of plans out there that use crock pots etc. for sous vide and that's a place to start. I like the idea that your circulating pump is on all the time but I cannot comment beyond that, a lot of pro equipment uses something similar with a thermistor probe and circulating pump so I think you're on the right track. Build one and try it.

I have done two DIY baths similar to what you are doing. (If you search the forums, you'll find a thread describing my first one in detail.) You can get it done for less than $50 with parts ordered from E*ay or Ama*on. You probably won't electrocute yourself, but I do strongly recommend you have someone knowledgeable in circuits (an electrician, or an electrical engineer) look at your results, not just this schematic, to determine whether it has been safely built. It is easy for a novice to make assembly mistakes (wrong materials, failure to properly ground, missing fuses, wrong wire sizes, etc.). This particular device needs that scrutiny, because you are mixing 120V AC with a tank of water, into which you'll be dipping your hands.

After building two of them, and dealing with the various problems that always occur with such things, I decided for my third one to just buy a simple Anova Sous Vide heater/circulator. It does exactly what my DIY ones did, but with much higher reliability and convenience. I bought mine with a clear plastic rectangular container to hold the water bath itself. If you buy the Anova on Amazon, they usually recommend such a clear container, and the Anova happens to attach well to it.

To others commenting on the use of a PID controller, a relay, and a 1500W heater: All are perfectly fine choices. The PID controller does just expect to operate a relay. The "P", "I", and "D" in PID is about the algorithm used internally to regulate, not how it communicates to the heater. The PID controller will automatically adjust to the performance of the 1500W heater, so it just means it will turn it on for shorter periods than lesser heaters.

the most important thing... use a ground fault outlet on both sides to be safe.

Damn... Another DIY project sacrificed at the altar of progress...

I have looked at the specs of the Anova. Quite impressive : 25°C to 99°C ± 0.01°C. The only problem : its too hot for black and white. The Sansaire is more versatile: 32°F - 212°F / 0°C - 100°C (± 0.1 degree at 140°F / 60°C). Precision at 38°C should be ok (C-41). I will look for other companies.

They do come with a price... But considering the convenience of these machines, I have decided to start saving... I have plenty of other DIY projects in the waiting.

I would like to thank everyone for your time and answers. Much appreciated.

R.

I have found an interesting alternative for people willing to do some DIY : Sainsmart RC200A temperature controllers (+/- 20$). They have a resolution and a precision of 0,1°C. This said, the DS18B20 waterproof sensor has a precision of 0,5 °C (which is good for C-41). All you need is a heating element. This can be a way for a precise water bath for less than 35$ (depending on your heating element). Way cheaper than a sous-vide pod.

Here is the link for the curious ones : Dead Link Removed

Dead Link Removed

R.

Sweet!

PIDs are nice for very precise control in a highly variable environment. They come in a number of forms, and yours is fairly typical. The relay is not unusual, especially if it is a solid state relay that can act quickly and frequently. Look carefully at the time constants involved so that you will be switching no more than about once a second. Also try for zero point switching to reduce RFI and improve the life of the components.

But all this may be overkill. I built a very successful temperature controlled warming bath using a deep plastic dishpan and a fully immersible fish tank heater. Adjusting the temperature is a pain, but once it is set it stays set, and I only use a single temperature. Other adjustable immersion heaters are available commercially, and they would probably be better than a fish tank heater (although a lot more expensive).

But then there is the hobby aspect. Designing and building a really good temperature controller would be a lot of fun and very rewarding.