Fermentor temperature controller

[WortMonger]

OK, now I am confused about the issue of copper and stainless (304). After reading up a little more on passive stainless and copper I am thinking it may be ok. Here's some of the info I got, the rest I didn't write down and can't find again (I am natorious for that). Here's one RadicalEd http://www.steelheadfasteners.com/acq.htm

Now I get to hear what virtually non-existant means, lol. I learn though so it is all good. I did some research on gold and chrome plating too just in case it is still a no-go for the copper and stainless :rockin: .

RadicalEd Question: If plating would chrome be better? I know that passive 304 stainless is actually passivated by a thin film of chromium oxide, so would chrome be a better fit? You see what I am saying chrome + chrome = no corrosion since they are the same metals, right? Dang I hope you have some good answers so I can learn more.

 

[RadicalEd]

I know that Stainless is pretty corrosion resistant, it's just a question of how much so. It very well might work, but before I was to conclude that I'd want to see some real-world data. But that I mean either a previous study conducted by a brewery, or a DIY experiment wherein you put something stainless and something copper in a covered dish of beer. If there's no corrosion within a month, it should be OK. Otherwise, it's back to the drawing board . Conicals are just to expensive to risk on this, though.

Gold and silver are just too darn expensive, so if you decide to plate it looks like your best options would be either chrome or nickel. I agree that chrome would likely be the best, but nickel should work just fine if chrome is just too much money.

Hope that helps!

 

[WortMonger]

Cool, well I will do more research on copper/ss and nickel/ss then. This is becoming more fun by the minute. I am finding out so much about copper that I never knew. Sorry, it has always been my favorite metal and now I might get to do a cool project with it. I am saving up pipe now to do a couple of oak leaves for the country club we belong to and their new building. It will be a much larger pour than my proposed cooling probe (I know you said concentric pipes would be better) project and so should give me ample practice in a first time investment casting. On the probe idea - forgot to ask earlier - what about a a thin diameter rod with larger "fins" sticking out from it. I ask because I know nothing about thermodynamics. I am just curious as to what say a 1/2" diameter copper rod cast to a copper plate would be remotely efficient?

 

[WortMonger]

OK, I called a local nickel plater and he said it wouldn't be that expensive for what I am wanting, nickel has shot to $20+ per pound now but he says a thousandth of an inch thick should do nicely and last forever. Now I just need more free copper, lol. Also, I was wondering about peltier size and places to buy them? Hope everyone is enjoying this thread as much as I am.

 

[Shockerengr]

Look into heat pipes....they are way more efficient at transferin heat in this sort of situation that just a rod would be.....and they use way less metal, albeit at an increase in manufacturing cost


having played with one, i can say they move heat impressively fast (held one at room temp, stuck the other end in hot water and couldn't hold onto more than a couple seconds before the whole thing was hot)

 

[RadicalEd]

Heat pipes are indeed the best way of moving heat, but I didn't mention them because I am not familiar with their effective temperature range. The way heat pipe work (for those unfamiliar with them) is a liquid within evaporates at the hot end, absorbing heat, and then condenses at the cool end, radiating heat. So the liquid has to boil at a temperature lower than what you're trying to cool. Ie if the liquid is water, the heat source has to be at 100*C to be effective. Of course, ones with a lower cooling point are available; otherwise they wouldn't be so wildly successful in PC cooling solutions.

You'd probably have to get one custom made, and it could get pricey. But good luck if you want to pursue it!

As for the pelts, check out your favorite computer enthusiast forum and read some of the stickies. www.extremeoverclocking.com and www.xtremesystems.org/forums are both excellent resources.

Best place to buy TEC's that I've seen is ebay.

 

[WortMonger]

Thanks for the info Ed. I have kept reading and think the mode of action for my idea at this point is to try the nickel plated copper heat sink (I'm thinking just like hot-side sinks will be only way taller to go into the fermenting beer to the bottom) and have some other brewing things plated at the same time. Should be under $100 and will look great. I mean I love copper but you do have to clean oxidation off all the time. I know that on the galvanic chart I definitely have less than a .15V difference with the passive stainless and the nickel. I know this kinda throws out being done for resale, as it raises the price significantly. I just want to make the most efficient one I can and replace TEC's if they go in the future cheaply. I see it as an acceptable/but high initial cost that allows a nice looking freestanding fermenter, without having to buy a cooling jacket or something that is expensive to ship from morebeer.com. I also like to show people my brewery and tell them I built it all.:rockin:

 

[nealmc]

I'm happy you guys are working on this, as I have played with a few TEC's as well and they do seem to have some potential given a proper power supply and thermal connection to the fermenting wort.

One thing that I haven't seen mentioned in your capacity calculations is the effect that the ambient temperature plays. I, for one, only have an air conditioner in my bedroom in my apartment and would love to ferment somewhere else, but during the summer, ambient temperatures in other parts of my apartment are typically around 80F. I know I could get a dorm fridge and it would work anywhere, but would the TEC devices you guys are working on be able to extract the heat from fermentation AND keep the wort several degrees below ambient?

 

[RadicalEd]

Glad to see that the ideas are coming along, wortmonger!

nealmc: what the TEC's can do depends entirely on their specs, and your PSU. The most important thing we need to find out is what voltage the TEC's run at, and supply a proper power source for it. A massive 320W TEC made to run at 24V isn't going to move much heat if you only supply 12V; a TEC designed for 12V would be much more efficient. The second most important factor is the Qmax; the amount of heat the unit can move (in watts) if there is no temperature difference across the hot side and cold side. What you want is a TEC with a Qmax higher than your estimated load; the 'extra' power in the TEC brings down the temperature on the cold side.

Check out the Peltier sections of the sites I linked to above; they explain it all much better than I can .

 

[nealmc]

Thanks for the links, RadicalEd. The only TEC's that I have played with have been rated at 12V, and I have tried several from I think around 320W to 120W. I was using a 12V supply and the biggest factor in the heat removal proved to be the current running through the TEC's. I eventually got a PC power supply capable of 25A at 12V and I can only let it run though the TEC for a couple seconds before the temperature at the hot side gets too close to Tmax. I don't have a huge heat sink like Yuri's, but I think you'd need at least something like his to let it run for longer.

Not being an electronics whiz, is there a cheap way to limit curent in a circuit (0-25A)? You could probably just turn it off and on for varying periods and the thermal resistance in the sysytem would even out the temperature cycling, but it sure would be nice to run a TEC cooler with a cheap current limiter hooked up to a PID controller or equivalent circuit.

 

[RadicalEd]

Can you limit just current? The short answer is no. But you can limit voltage, which in turn limits the current. TEC's have an effective resistance, and as such must abide by the ironclad rule of I=V/R. With R holding steady, a drop in volts should mean a proportional drop in current.

I personally suspect that the best method for controlling such high currents would be pulse width modulation, which is pretty much what you were thinking. A PWM device pulses the voltage across it at varying intervals; put a nice big capacitor on the TEC side and the voltage will nicely average out the pulses into a relatively constant voltage. Make the pulses longer, and the average increases, and vice versa. You may need a relay to handle the currents, though. An extra advantage of this is that PID control is very easy to implement. Looks like Wikipedia has some good stuff, check it out.