Sous Vide anyone ?

[wheels]

Thanks Jamie. I think that I get away with using the STC-1000 as I use the slow cooker on the 'keep warm' setting. The heat is gentle enough to give time for the thermostat to respond. It seems to work fine for steaks and such like, but I wouldn't want to use it for more 'temperature sensitive' projects.

Anyone got an 'idiots guide' to setting up, and working with, the type of PID controllers on Ebay?

Phil

[Jabba]

Sorry Pill, I've not set up one an off-the-shelf PID controllers, but I know many home brewers have used them, a lot of info should be out there on the 'net. The sum total of my PID education has come from the chap that wrote the PID 'library' of code for the Arduino, and most of that goes over my head!

Oh, here's some documentation relating to another implementation of the Arduino library.

Cheers,
Jamie

[Dingo]

I think, i could be wrong, that there needs to be a bit of clarification on what PID actually is. Look away if you already understand. Or tear me up if i'm wrong.

The easiest way to understand PID is the cruise control in your car. It is continually measuring and adjusting your speed and comparing it to a set point. Actual speed outside the set point is considered an "error" and the system strives to correct that error, thus achieving constant speed.

For our purposes, let's consider trying to achieve a sous vide temp (or constant speed) of 130F; and here is the fundamental difference between a temp controller and true PID;

A simple temp controller (which is considered digital..i.e. on/off function) will switch a heating element on or off, within a "range". That range could be anything, depending on the accuracy/quality of the controller. For arguments sake let's say +/- 5 degrees. If the set point is 130F...the controller will switch on at 125F and switch off at 130F. (in a perfect world)

A true PID system (Considered "Analogue") will control the voltage to the heating element to achieve the target. i.e. the set point. What does that mean in terms of temp? Depending on the programming, in a heat cycle it will vary voltage to the heat element to maintain the set point. As the difference in temp, or "error" is increased, so to the voltage/temp is increased.

So, back to the cruise control analogy...if we had a digital cruise control with a +/- 5mph accuracy, our speed, at a set point of 100mph (i like it fast) would cycle between 95 and 105mph. Where as our actual cruise controls are far more accurate being that they are true PID.

What does all this rambling mean to us? In most cases, a standard and less complicated digital solution is fine for our purposes. i.e. curing cabinets etc. I'm guessing that even for sous vide a high quality low variance temp controller ( +/- 1F) would do the job.

I'm not knocking PID as a solution, if you really want to dial your temp control in. However for the cost/complexity involved i think that for 99% of our needs...Keep It Simple Stupid

[DiggingDogFarm]

I believe in keeping things simple but there are a few situation where a PID can help....when cooking things that are very temperature sensitive (such as eggs)...it's also important to have the best control, accuracy and precision at lower sous vide temperatures where safety is even more of an issue than normal.

Most of the cheap PIDs that are commonly available aren't all that accurate and precise for sous vide.

The Sous Vide Magic is what I currently have...it's accurate, precise and a programmable PID.

http://freshmealssolutions.com/index.ph ... mid=100086

He ships just about anywhere.


~Martin

[ped]

This is all very useful info thank you, just a question DDF, what do you use the controller in conjunction with, and is there a need for the pump that Phil suggested?

[wheels]

I accept DiggingDogFarm's points regarding accuracy and low temperatures, but I'm just not seeing the fluctuations that are being mentioned elsewhere. I think that with using that with using a slow cooker that only has a small element which is very slow to respond, the see-sawing of the temperature seems to be avoided. When set to keep warm, the (very small) element is only trying to achieve 72°C (ish), so when on will only increase temperature at about 1° per 15 - 30 minutes or so. This gives time for the 'time-delay' built into the thermostat to have expired, and for the unit to switch the element off before the temperature has risen too much.

Or, at least, that's my take on what's happening. All I can say is that, within the caveats mentioned, it seems to work. That said, if I had the money...

Phil

[Wunderdave]

It's important to differentiate between a bang-bang temperature controller and a PID controller. A bang bang/on-off controller uses the setpoint calculation like dingo mentioned above. PID controllers use a mathematic algorithm to figure the response of the medium to the heat input provided. It is not my experience with a PID that the voltage is modulated, but only the time that the voltage is applied. So you will see a flicker on/off. Compare that to a bang bang where the voltage will stay on until the set point is reached, often causing overshoot.

So as your cooking medium (water) comes to temp, the computer chip in the PID measures the response curve. Then during the first few fluctuations, the controller measures how far the medium temperature fluctuates and adjustes the heat input to maintain that temperature steady. It's a reducing sine curve.

Depending on your tuning setup of the PID (the P, I, and especially D figures), it will look something like this


Some PID controllers are able to auto-tune to optimize, some you will need to tinker with the settings yourself.

[wheels]

Thanks, that illustrates the difference well.

Have any members got any suggestions for which PID to use? That is, one that will give the accuracy that DiggingDogFarm points out isn't present in the very cheap versions, but which doesn't break the bank. Oh, and either available in the UK, or that can be shipped to, and used, in the UK.

Phil

[DiggingDogFarm]

ped,

I have the Fresh Meal Magic heating unit from the same supplier.
The controller can also be used in any non-digital heating vessel....preferably with a bottom element, such as an electric coffee urn.

Circulation is definitely a good idea because natural convection isn't reliabe unless the unit is ramping up to temperature....when maintaining temperature convection currents are minimal.

I use a simple air bubbler.

HTH

~Martin

[Dingo]

Yoyo, sorry for the long post. DONT READ UNLESS YOU ARE INTERESTED IN PID or want to argue for fun

Thanks, that illustrates the difference well.

Have any members got any suggestions for which PID to use? That is, one that will give the accuracy that DiggingDogFarm points out isn't present in the very cheap versions, but which doesn't break the bank. Oh, and either available in the UK, or that can be shipped to, and used, in the UK.

Phil

For simplicity and cost...buy this;

Re: Sous Vide anyone ?

Postby quietwatersfarm » Tue Dec 17, 2013 1:03 pm
Just to say that these http://anovaculinary.com/ are the business. Wouldnt swap for anything, running three units here in big plastic cambro tubs.

Seriously, unless you've got cash, time and nothing else to do..it is the way That's what I'm gonna do!

I used to work for these guys;

http://www.mtl-inst.com/

One of the things they specialize in is process control for the hazardous materials industry.

I'm gonna go in into more detail (sorry )...

So if you DIY guys are looking for the easy way out...see above. If you are really bored and have nothing else to read...go on....

Okay..lets start by understanding how a thermostat works... For simplicity, it is two different types of metal strips glued together. Each metal expands at a different rate according to temperature. This causes the strip to bend...and press on a switch. On or Off. That's it. Depending on the metals used, the range, and/or rate/accuracy of the switch will vary. However, for our purposes...it switches on or off. This is considered digital and can be considered I or O. The dial you turn just pre-loads the strip.

Now for the PID quick lesson. The definition of PID is here;
http://en.wikipedia.org/wiki/PID_controller

What does that mean for us meat makers?.. And to the point Sous Vide?

In practicality, most variables (ie temp, flow, PSI etc) can be measured two ways. As an analogue variable or a digital variable. A digital variable describes a state, either on or off, I or O. Or an analogue variable describes the condition. For example.... a pump switches on....

Digital.....the pump is on
Analogue... the pump is flowing 10 gpm

And thus analogue was born. Analogue measures a range, digital monitors a condition. Both have a roll in true PID. How does analogue work? The most common version is 4-20mA. i.e a temperature sensor is set to measure between 0-100F...the sensor will out put 4mA for 0F and 20mA for 100F. The controller will see the same.

So, the same can be said for an output...Digtial being...on or off, or analogue can be a variable. For the same example...

Digital...turn pump on or off
Analogue...run pump at 8gpm

Okay, cool, we figured out how to measure and control things. What know? Well, the second part of the renowned "PID" solution is the controller. Wonderdave mentioned controllers...well he is 100% correct. In our market they come in many forms...the most common are; "Smart Relay" and a "PLC" ...(Programmable logic controller). What they do is interpret the data from the sensors. Either Digital or Analogue and apply the "algorithm" that Wonderdave mentioned, and apply an output. PLC's ....you need to be a rocket scientist to program (not actually true..but pretty close)...Smart relays....a lot easier to program these days with free software available.

So..taking our pump example.... the pump can flow anything from 0-100gpm...our set point is 100gpm....the controller turns the pump on and assesses the flow (via algorithms and 4-20ma sensor). Initially it sees the flow as 0 gpm (4mA as the pump hasn't started), so in an ideal world speeds the pump up to the set point of 100gpm (20mA). So if we wanted a 50gpm flow...we would set the controller for 12mA (16mA availble for 100gpm span) and the controller would slow the pump speed down to 50% or 12mA on the sensor.
The behavior of the controller and the pump are a function of the software in the PLC. Depending on the function...pump speed, temp, power factor etc might need to be controlled also. And thus to my point....all of this conjecture, and boring reading, could be avoided by purchasing a ready made sous vide for a couple of hundred bucks....or you can undertake a process engineering degree and do it your self.

I know this was long winded...but the degree is many years. I'm not an engineer by any means...but I've played, and am still playing, with PID solutions every day. If you have a question let me know. I probably wont know the answer, but i will figure it out.

[Jabba]

Very informative post, Dingo! I would however not agree with your point about needing a process engineering degree to build a DIY version! As I mentioned above there's a PID library for the Arduino (a microprocessing platform designed for artists and hobbyists- not engineers) that is very easy to use- all the hard work has been done in the background- and has been used in a number of sous vide Kickstarter products.

I got into Ardunio when I built my home brewery- I no engineer but by looking at the plethora of examples and forums on the 'net I soon picked it up. It's easy to get started, I bought a starter kit for £35ish. Its great fun, sometimes frustrating, but ultimately very satisfying building your own computer controlled piece of kit!

Cheers,
Jamie

[wheels]

The unit posted by QWF, or any at a similar price, are out of the question for me; I just can't justify spending that amount on it.

To be honest, I'm quite satisfied with my cheapo Chinese temperature controller, and it does what I've asked of it so far. However, I was wondering whether I could have an even better set up by using something like this:

http://www.ebay.co.uk/itm/4-IN-1-PID-KI ... 1235547540?

I know, it's not as good quality, or as reliable. But would it be better than the other cheapo controller (STC-1000) that it would replace?

Phil

[Jabba]

That appears to be the SSR and heatsink I have, it's been reliable so far! With the PID controller, whilst maybe not the quality of an Auber unit, it appears to be a good deal. I'd have a punt on that!

Cheers,
Jamie

[Wunderdave]

All this chatter inspired me to open the box of the controller I built to see if I could figure out what's wrong. Loose connection found & soldered; no more looseness anticipated. Now I need to find some inspiration for cooking...

Thanks guys.

[saucisson]

I did my ham on the stove the old fashioned way in the end because I discovered my vac bags weren't suitable for "boil in the bag/sous vide" I've ordered in some different ones.