PID Controller

I'm not sure of your calculations but most controllers to a greater or lesser extent have a temp hysteresis. Whether this is an issue depends on what you're cooking and the temps involved. The fairly large hysteresis you calculate will mean you can't achieve the subtle variations on sensitive proteins such as eggs and maybe not fish/seafood but probably will be ok for long cooking of meat. You shouldn't cook food at very low temps because of temp uncertainty and bacteriological activity.

I've used an Ebay PID to control my Bradley smoker but you then need a few other components to be able to connect everything together. Auber Instruments do a plug and play type solution:

http://www.auberins.com/index.php?main_page=index&cPath=13&zenid=ecafebee0294c7ce8f57f01e0aeb8e11

Thanks. Anyway I assume that as long as I stay above 131F a couple of degrees variations above that should pose any safety issue!

Re PID: I've been looking around for DIY guides but found very little. I'd go for the Auber or Sousvide magic but they are expensive.....it's all starting to feel like a lot of work for a steak!

Have you seen this DIY project by any chance?

http://seattlefoodgeek.com/2010/02/diy-sous-vide-heating-immersion-circulator-for-about-75/

Thanks! Anyway I'm already tired of sous vide, too many cables, kitchen looks like a lab! and food comes out in the shape of the plastic bag......naaah, I'm back to my pots and pans, at least for now!....might save money for a SousVide Supreme for next year though!

I use a 7 quart oval CrockPot (Rival SCV700B) and a Johnson Controls A419 thermostat for most sous vide cooking. The thermostat has a programmable hysteresis, but it cannot be set to less than 1 degree. There's overshoot and undershoot. It's not as severe as your experience, but clearly the water temperature oscillates. I also have a PID controller that can be used with the CrockPot, but I generally use the thermostat.

Here's why:

First, I experimented with a meat thermometer in a 2 lb beef roast. I discovered that the water temperature oscillates rapidly enough that the temperature variations don't penetrate the roast. An eighth of an inch inside of the roast the temperature is rock steady. The results are consistently perfect. There's no greyness layer on the outside--perfect pink color right out to the thin seared layer.

Second, the thermostat recovers from short power outages without a hiccup. In my neighborhood, we get short power down transients probably due to switching about once a month. They last perhaps a quarter of a second, but that's enough to reset the clock on the microwave, cause the cable box to reset, shut down computers, and upset the PID controller. If I am at work when this happens, the meal will spoil before I get home with the PID controller, but the thermostat has no such problem.

Another advantage to the thermostat is that it brings the water bath temperature up to target after the shock of adding refrigerator temperature meat much faster that the PID controller. The PID controller goes proportional and very slowly approaches that 131 degree (safe zone) temperature. This affects things like cooking time.

So the amount of temperature overshoot and undershoot I get with the thermostat has no observable disadvantages. You seem to be describing a bit more overshoot/undershoot than I observe with my own equipment.

By the way, I tend to prefer using degrees F in lieu of degrees C. That makes my hysteresis of 1 degree F and overshoot/ undershoot significantly less than your own observed...