DIY Sous Vide Controller

[KennethT]

There is a HUGE amount of discussion and info about doing this in the old sous vide thread... the thread was so big, they made an index for it - go through the index and find the stuff about DIY rigs... there's more info in the new SV thread as well...

[Robert Jueneman]

Lightobject makes an inexpensive dual temperature PID controller for only $38.50. It's available on Amazon at http://www.amazon.com/s/ref=nb_sb_noss?url=search-alias%3Daps&field-keywords=sous+vide+contoller&x=0&y=0

Another $3.95 for a thermocouple, plus $8.50 for a 20 amp solid state relay, plus $15.50 for a 100C DC water pump that is FDA approved to circulate the water. Add a Marshaltown 742g Bucket Water heater for $39.99, and maybe a Cambro tank, and you’ve got a pretty nice DIY sous vide machine for not much more than $100. Of course it won't be pretty, and you are pretty much on your own in figuring out how to program a PID controller. But the circulating pump should minimize the overshoot problems that a conventional PID controller in a non-circulating bath has to cope with.

I have to believe that AllClad and others are beginning to eye this market as well. It really isn't all that hard.

[paulpegg]

I work in the computer industry and recently saw an interesting project on a BBQ forum that I frequent that took a retail Internet router and made it into a automatic temperature controller (ATC) for smokers - LinkMeter

Using that device as a starting point, I would think it should be fairly to make a Sous Vide Controller. All you would need is controllable heater and circulation pump. Are there people doing these types of DIY projects and if so, what types of controllable heaters and pumps would they recommend.

-- Mache

I posted this MY SOUS VIDE COOKER on the now closed sous vide topic. Since I wrote this i have upgraded the pump to a high temperature centrifugal pump with silicon rubber hoses. I use it every week and it works beautifully. There is also a link to PID Controller settings which are set up for a CD101. i found that these settings are pretty common and will work on other models as well.

Edited August 16, 2011 by paulpegg (log)

[Brasshopper]

Absolutely. I wrote one of those Amazon Guides and a "listmania" list that gives links to all the parts you need to make one on Amazon. Seattlefoodgeek has the canonical set of instructions, but I had a hard time with his approach until I realized that the whole idea of building an immersion circulator is wrongheaded (for me, anyway) and way too complex, while at the same time being not reliable enough because of the complexity. If you look at the appliance world - from manual rice cookers, to slow cookers, to tabletop roasters, to hot plates, to electric frying pans pretty much everyone has a device that can heat water and can be controlled by feeding it power or not. Plug the device into a straightforward PID controller hooked to a Pt100 and an SSR, and the whole thing can be done for under $100 in parts. I had two roasters and an old rice cooker that could be controlled as well as an electric pressure cooker and an electric slow cooker that could not be controlled. So three of my five appliances are OK for sous vide.

There are people who don't want to do the wiring but many do - I built two of the devices. The "Guide" thing is here and The "Listmania" thing is here.. The guide is the more complete writeup, there is a different in the allowed format so I had to do the writeups differently. Both are at the maximum length Amazon allows for such a list.

When I built mine, I didn't know whether the stuff I wanted was exactly what the stuff on the internet was talking about and ended up trusting Amazon's "people who buy X frequently also buy Y."

I guess that the quantity 1 retail on a device that does what you need to do for sous vide regarding tight temperature control, the PID, is about $39, and it needs to control a SSR to provide the 110 volt power control. That device is about $10. If you actually plan on controlling a hot plate or something that runs up around 1800 watts, you will want the heat sink. And you need a thermocouple or, better, a Pt100 temperature probe that is OK for water.

Yep, that is more than an old router, but you get your display, it is all done. If the router was still on wi-fi so that you could log in to it, get it to do strip chart style recording and stuff, that would be cool...and it might be worth just adding the probes to one of these so that you could record how well you were holding temperature...

Now the lowest cost scheme I know of is that the same PID I point to has an internal relay that can handle 300 watts (which is very close to its capacity) but many of those immersion heaters that can be used to heat a coffee cup of water to boiling are 300 watts. I have seen plans that uses the cord from the 300 watt immersion circulator to provide power to the PID, and then the PID's internal relay is used to power the immersion heater - so, literally, the whole thing is done for less than $50 in parts - you don't even have a second line cord. I thought that was a bit skimpy. Recycled chopsticks are used to hold the probe and heater in the water :-)

In terms of whether you need circulation for the device, well, there are several schools of thought. If your device heats the water from the bottom, like a rice cooker, you probably do not need a circulating pump. If your device heats from the side, like a slow cooker or a roaster, you probably do.

There are two sizes of circulating pumps: Too small, and too big.

I have 2 working pumps and one melted pump which I keep to remind me that I should be careful and not let the thermocouple fall out of the water.

One of the pumps I have is the 6 volt 100C pump that lightobject sells - it is a 1 liter per minute pump and is in the "too small" category, I use it to stir a small rice cooker that I use for a pork chop or single bag of vegetables. I tried it in the roasters, and you can't feel the circulation at the other end of the pan, which makes me think that it is not large enough. I use it with a "many different voltage" brick that I got off the net, and I found that the voltages on the slide switch were printed on April 1. Many people complain that the power supply overheats. Of course it does, when the device is being run with twice the voltage it is supposed to have it is pushing twice the watts. It produces 4.9 volts, perfect for this pump, at the 3 volt setting - and it does not change the voltage significantly when the load is attached.

There are people using heating elements that sink into a bucket or a pot, and those frequently use an aquarium air pump to stir with bubbles. There are a bunch of different pumps that are used for those little fountains, and they work great if the water is not too hot - one run at 180F will break them, and they work fine at about 150F, and somewhere between 150F and 180F they melt just enough to fail. Lightobject has pumps that can handle 100C but not immersed, since they are pretty much always used immersed in this application, they are less than useful. You can get food grade silicone tubing that is the same size as aquarium air hose - you can plug the end with a 1/2" stainless nut by folding the end over, and a couple holes in the side of the hose makes a bubbler which stirs real well.

[dcarch]

For a water heater, you can try using a 220vac 2,000 watt heater and run it at 110vac.

If you still want to run it at even lower wattage, run it thru a diode which will cut the current (watts) in half again.

Any time you run an electric heating element at lower wattage, you extend the useful life of the heater greatly.

dcarch

[Paul Kierstead]

Just what is the normal useful life of a immersion heater? In my very casual observation, it is on the order of a decade+ assuming you don't run it dry.

[dcarch]

Just what is the normal useful life of a immersion heater? In my very casual observation, it is on the order of a decade+ assuming you don't run it dry.

I don't know. Never tested it.

However, they do sell 130v light bulbs to run on 120v for areas where you don't want to replace bulbs frequently. 130V bulbs running on 120V lasts 2.8 times longer.

A resistence heater is the same as a light bulb. So running a 220v heater on 110 v, the heater probably will last 10,000 years.

dcarch

[Paul Kierstead]

Well, a heater isn't really on the brink of burning up like a light bulb is. But we digress

[dougal]

Really, its much simpler (and thus quicker) to use a cheap eBay PID controller module (with autotune) rather than to build and program such a thing yourself, even starting with the Arduino platform.

If you aim for a heated waterbath with base heat (in the style of the Sous Vide Supreme products) rather than a circulator that can be used in any pan, you make things much simpler for yourself.

A non-shallow bath with well-spread base heat doesn't need constant stirring/agitation/pumping.

The modular approach leads people to buy rice cookers and such -I have a 7 US Gallon tea-urn {water boiler} - as the bath + heater element (thereby taking out a lot of mains electricity + water safety/sealing/robustness concerns).

Naturally, its important that such a unit be completely dumb, totally lacking its own electronic smarts. A crude mechanical thermostat is no bad thing, since it can be set a little high and used as an over-temperature safety cutout.

I'd suggest you find a water-heating-vessel, a Solid State Relay (eBay again?) and a PID module that both works with your choice of immersible temperature probe and has an SSR drive output.

Go with that and get cooking.

THEN, if you have partitioned the hardware appropriately, you could swap in your homebrew Arduino-based controller to command the SSR in place of the bought-in module anytime you wanted to experiment.

One possible benefit to a homebrew design could be the ability to use multiple temperature sensors, distributed around/throughout the bath, and have the controller 'intelligently' deal with non-uniformity in the bath.

However, since the optimum PID terms are going to be different for stirred/unstirred conditions, this could be creating a significant control problem. But maybe that's what you are looking for?

Edited August 18, 2011 by dougal (log)

[HowardLi]

OK - before the summer is over I will be trying to get a smoke job done on some pork with this unit. I've happened on a kind-of-but-not-really-reasonably-priced film RTD at McMaster-Carr, PN 6568T46. 3-wire 100 ohm; most likely European standard tempco. I'll attach it to something thin and put it inside a cardboard box, with a wood chip-filled cast iron pan on the hot plate doing the heating.

[HowardLi]

I got the probe thermometer. It's got a ring terminal for mounting so I'm going to attach it to some aluminum square tubing for mechanical and temperature stability.

I also waterproofed my RTD by slipping in some silicone tubing up to the hex nut and crimping it. It's not a perfect solution, but now water won't get inside. I had it going right into the water before and it caused the RTD to intermittently fail, which was (obviously) bothering me.

Pics to come soon.

[HowardLi]

RTD in standard configuration with coil relief

Relief removed

Silicone tubing slipped over and clamped