Wednesday, November 4, 2009

Home made electric smoker with PID feedback temperature control

updated again 9/17/12 (at the bottom)

I know this isn't the newest idea on the block (heck, Alton Brown made one of these things), but I thought I'd post a writeup of my home-made electric smoker with PID temperature control, because it's slightly different from the standard version.

WARNINGS:
Before I get started, I'd like to offer a couple of warnings.
This project requires messing around with 115v mains current.
This project involves VERY hot items, and a definite possibility of fires, do NOT operate this thing indoors, or anywhere flammable.
I assume no responsibility whatsoever for anything you may do or that may happen as a result of your attempts to follow this build process


MATERIALS:
  • 2 large ceramic planters
Due to some experimenting that resulted in breakage, i ended up with one planter slightly smaller than the other. In the end, this works out well, as they nest together pretty tightly.

  • hot plate
I got a crappy one from target for about $20. I have heard that the ones with a normal stove-element looking coil are better, but mine has an element contained within a flat casting, and it seems to work fine

  • bricks
About 4 bricks. 3 of them go under the bottom planter to lift it up off the ground a bit and give me access to the hole, and one is split in half supporting the hot plate inside

  • grate
Originally we had a round grill grate that i found at Home Depot, but i seem to have lost it in the move. For the last couple of smokes I've been using a cheap drying rack sort of a thing that I bent to make fit. As long as it doesn't collapse, it's really not particularly important. The drying rack's rectangular nature does make it easier to toss a few more chips into the pan though

  • chip pan
I tried using a stainless stel mixing bowl, but the bottom gave up after only one or 2 smokes. They really aren't built to take this kind of abuse. I replaced it with a pretty solid skillet-like pan (avoid teflon, this thing gets well above the temperatures where teflon can give you problems). It has a very heavy reinforced bottom, which seems to be taking the abuse well.

  • water dish
I'm not 100% sure on the necessity of this, and I don't always use it. But anyway, it's a small stainless mixing bowl that i set right inside the chip pan.

  • PID temperature controller and thermocouple
Bought for about $35 off of ebay. It's a very common piece of process control equipment. Mine came with a K-type general purpose thermocouple.

  • Solid state relay
My PID can't switch the ~7+ amps of 115v AC that the hot plate draws, so i need to have it trigger a relay that can. I chose a solid state relay because it's silent and can handle rapid switching. I could probably have got by with a cheaper mechanical relay, but to be hones, I already had the SSR, so I used it.

  • enclosure
I got lucky and found a sort of outdoor outlet box type thing at Home Depot (actually, Josh found it), and it is JUST the right size, as well as being quite robust and weather proof (or at least it wil be once I finish sealing up the holes I drilled in it). The enclosure is fairly important, as you've got live mains power running in and out of this thing. You really don't want to touch the wrong part.

  • high amperage quick connector
I don't like to leave this whole thing hooked up, so I needed an easy way to put the electronics away without having to take apart the whole smoker. I picked up a 10 amp-rated 2-conductor quick disconnect type thing at you do it, and it has served its purpose well. Although due to the way it's set up it can shock you if you plug the whole system in before you connect it to the hot plate. this definitely needs to be addressed. But so far only Josh has been zapped.

That's all you NEED to make this run (need is a questionable word. you can do all of this without the PID or anything like that, but that's boring) . The following items are things that I've added that make things easier to deal with.

  • remote meat thermometer
You need to know the internal temp of the meat. The easiest way to do that is to put a probe in it, run it up through the hole in the top, and connect it to it's little display piece. it works well

  • datalogging
I have an onset u12 datalogger with 2 remote temp probes set up. The ability to go back and see what was going on is invaluable in terms of improving your process! In the future I'd like all of the monitoring and logging to be handled by my laptop via Arduino, and to be web viewable, but I'm not quite there yet.

CONSTRUCTION:

Most of this is very straightforward. Set the bottom planter on some bricks, pull the heating element out of the hot plate and put it on some brick pieces inside the bottom planter. Try to make it nice and stable. My hot plate had some temp control electronics (a simple bi-metallic strip) and a lamp attached to it. I ditched all that, but kept the fiber-coated heat resistant wiring. I ran the hi-temp wires from the hot plate element through the bottom hole, where I soldered them to some standard heavy wire to form the hot plate power cord. The chip pan goes straight on top of the hot plate element, the grate wedges in to the bottom planter, and the top planter just sits right on top.

At this point, you could just plug the hot plate in, and the whole thing would get VERY hot (350-400*F at least), and you'd probably burn out the hot plate element after a while. What you need is some sort of temperature control. The hot plate likely came with a simple bi metallic strip and knob. The problem with using these is that there's really no simple, effective way to do it. If you put the knob/strip outside the planter, it never gets hot and doesn't do anything. If you put it inside, it's impossible to adjust without largely disassembling everything and probably burning all the hair off your hand. Further, if you move the strip physically a few inches toward or away from the element, your temp control completely changes. And finally, even if you can get it all nailed down, it's still complete trial and error trying to get the temperature set to where you want it, and you've got no way to adjust for temperature creeping up or down without physically being there to fiddle with it. This is where our friend the PID temperature controller comes in.

WIRING/PID HOOKUP:

The remaining construction involves hooking up the PID.

PID Power:
The PID runs off of standard wall current, so you could have just one plug that powers both the PID and the hot plate. For various reasons, I chose to have them separate. One plug powers up the PID, and a different one feeds power through the temp controller to the hot plate. This way I can have the PID on and mess with it's settings without having mains voltage elsewhere in the box. Anyway, run power to the appropriate terminals on teh back of the PID.

Thermocouple:
After power, you need to hook up the thermocouple. My PID came with a nice K-type thermocouple with about 4 feet of braided-steel-encased wire, which is perfect for my setup, but not long enough. You need to be slightly careful about extending the wire, as variations in resistance will change the temperature reading. I just used about 6 feet of half decent speaker wire soldered to the original thermocouple wire, and it seems to be fine. Hook the + and - thermocouple wires to the terminals at the back of the PID. Pay attention to the polarity. If you have it backwards, the temperature reading will go down as temp rises, and things will go quite badly awry.

Relay and Hotplate Power:
My PID has an internal SSR output, basically it puts out 12vdc when it wants to be heating, and 0vdc when it doesn't. This is perfect, because it can serve as a trigger signal to the SSR with no modification. The Chinese instructions that came with the PID didn't make this super clear, so I wasted some time and effort hooking up a 12vdc trigger power supply that I didn't end up needing. Anyway, wire the positive and negative SSR outputs from the PID to the trigger side of the relay. Take the cord you want to use for the hot plate power supply and split one of the conductors, attaching each end to the output side of the relay. Now when the relay closes, you get mains current flowing through the hot plate. When it opens, the hot plate shuts off. Be aware that any time you have this plugged in, you WILL have live mains current in the hotplate power wire, even if the PID is not calling for heat.

FINAL ASSEMBLY:
Box all of this up inside the enclosure, hook up your high-amperage connectors to the hot plate and the relay, and you're ready to test. My PID has an auto-tune feature, but it doesn't seem to work very well for this application. Maybe it will work better for you... I have achieved pretty good control results through trial and error with P=3, I=1, and D=1. Very basic PID control theory is discussed here: wikipedia.

For best results, put the thermocouple right near the grate surface, but not touching either the grate or the meat. Plug in your thermometers and your dataloggers and have at it!

Here are a couple data plots from smokes I have done. You can see that the PID does a very good job controlling the temperature inside the smoker, even if the outdoor temperature changes significantly. The major changes in logged temp are either from me opening the smoker to poke at things, or from me changing the PID set point.

This plot was from earlier, when I hadn't got the PID values settled down quite as well:


This plot is a later smoke, with better temp control:

So far this whole setup has been very safe and reliable. As I said, Josh got zapped by the connector once. It has never caught fire on me, but it certainly could, so I never operate it within about 10 feet of the house, just to be safe.


UPDATE: I ended up putting a second burner in next to the original one, and I can hit temps near 400*F now. The only additional electronics I needed to deal with for the upgrade was to order a higher-amp rated SSR, and install a heat sink for it (I tried it once without the better relay and heat sink, and the whole thing melted down).

UPDATE2: I also ended up cutting up a weber charcoal grill grate to make it fit, which works better. The smoker I made for my dad is larger, and a regular weber grill grate fits in it perfectly without modification (I think his uses 21 or 22" planters). I also switched to the coil-style heating elements I mentioned in the build log. They do seem to both heat faster and last longer. I scavenged them from a couple of $15 hot plates I bought off of amazon.

UPDATE3: Hackaday picked this up, so I thought I'd add a few more pics. Pardon the mess, I forgot to put the tarp over the smoker, and it's been raining.









9/17/12 update here's a little time-lapse vid I made of me making 4 smoker brains for family members:

8 comments:

  1. Your blog works! And you are a dork.

    :) Looks nice though

    ReplyDelete
  2. Hey there, great way of expanding on a simple idea! im looking to do a setup very similar to what you have, been collecting parts. now its time to order the heart of the electrical components, which i admit i am terrible with! so looking for a little help on picking out the right items to order. actually i created this google account just to post on this blog and try to get some feedback where i need help. is there somewhere i can email you and send some links to see if im on the right track with my electrical components? Thanks in advance for anymore help you can give!

    ReplyDelete
  3. Can you pose a couple more pictures of the enclosure you're using for the PIC controller?

    ReplyDelete
  4. Wow I did a double take on your PID enclosure, I used the exact same enclosure for my PID controller that holds my garage fridge at 68-70F for fermenting my beer! Your smoker controller and my fermenter fridge controller look identical next to each other.

    ReplyDelete
  5. Nice build, I have a similar setup (also shamelessly 'borrowed' from A.B.). One comment I have on thermocouples - your extension method can work ok, but generally the best way to extend a TC is to use the same type of wire as the TC. Otherwise, each joint (or junction) actually becomes its own TC, which can cause some issues with the measurement. In practice though, as long as the temp of each of these joints stays constant, it isn't that big of a deal. Keep on Smokin'!

    ReplyDelete
  6. Yes, you are correct. I forgot to mention that I replaced the jury rigged thermocouple with one that was long enough out of the box, and also nicely shielded with stainless steel braid.

    ReplyDelete
  7. This comment has been removed by the author.

    ReplyDelete
  8. Ide love to see a simple side view diagram of this, im about to move into a apt complex and would love to build one of these

    ReplyDelete