Everett's SC/TO Build

[Everett]

So if anyone is curious as to what I actually did for my roasting setup, it ended up being 2 SC/TOs. Just pulled my first 2 roasts from them on Monday night.

First, I started by separating the heating coil and stirring motor of both Stir Crazy's. I put them each on their own plug. Once they were all assembled, I threw 1# of green beans in there to test with my modified stirring arms, and at best, it gave me angel wings. Perfect...

At this point, I played around with modifying the arms slightly to see if I could get proper stirring. It was always negligible for the 1lb after roast I am shooting for with these guys, so I decided not to spare the expense, and to modify further. Father in law has a couple nice 60RPM electric motors sitting around (with about 100 other various ones), so I set to work with the brother in law. Got those big boys mounted in there, and it was a night and day difference. Very difficult to bog these down.

Next came the spacer issue. I used a 3/4" strip of aluminum, and bent it and fastened in 3 spots much like what Peter has done with his. I can push the overlap open to let the chaff out and let it spring back to close to retain heat. Simple enough.

After this, I needed to find somewhere to mount these, as they now have about 6-7" of motor hanging off the bottom, so I picked up a couple alarm panel boxes I had around, and set to work mounting the things to those. I figured I needed a spacer, and now there is about 1/4" clearance from the bottom of the motor to the bottom of the box. The box is hinged in the front, so I can lift it over and dump the beans. Seems to work beautifully. The plan is to mount the bottom part of the box to their final resting place and build a bean cooler that will sit in the center of them, with a chute going up to each side to pivot the SC on the hinged box when dumping.

Arm attachment was another whole ordeal with the new motors, as they are a standard 5/16" shaft. I had thought of grinding a slot in the top, and fastening the arm in there somehow, or getting some thick stainless piano wire, and spring tensioning the arm on it. In the end, I borrowed a die from my Dad, and threaded the shafts. The darn things were super hard. Was quite an ordeal, but it works beautifully. Then I threaded on a stainless nut, and put a crush washer in, next came the arm, and a final nut. All tightened together, it works wonderful. I did everything previously stated to both SC's using identical hardware and motor.

The next trick was to deal with the Turbo Ovens. (Sunpentown SO-2000). Started out the same as with the SC, with pulling them apart. First one I had trouble, and ended up trying to pull the fan off the shaft before I realized that there are 4 screws down in recessed holes, that hold the top plastic on. Took the top cover off, and set to rewiring this guy totally as well. Removed the handle switch, and the timer knob, and wired cords directly. One cord goes to the fan, as well as the Power LED I believe, so the LED is on when the fan is on. The fan cord, I used a grounded cord for, and put a ground on the fan. The other is wired directly to the heater, with the LED hooked up as it was stock. Modified the top case with the dremel a bit, so the other cord exits beside the stock cord. Reassembled the TO, and straightened the fan blades as best I could. Plugged in and tested, and the fan and heat now work independently, but the heat knob still needs to be set to full. Will be using router controllers to control the heating coils.

The next step, now that both parts were working as they should, was a control panel for said devices. A picture is worth a thousand works here. Basically, it is a double box with 2 receptacles, all separated. Below is a 3 gang box with 3 switches that control the receptacles from top left, top right, to bottom left. The bottom right is hot all the time, and that is where the router control is plugged in. I found this to be the most cost effective option, as single switched and single gangable boxes are super cheap. The special double switches are expensive, and I don't need one for the router control. It's all mounted on some identically sized offcuts of melamine I had sitting by the saw.

[size=78%]





[/size]

[Tex]

Good looking setup! Only things missing are a Kill-A-Watt and digital thermometer probe?

[Everett]

Kill-a-Watt would be handy. I do need to get a couple eventually, but those will likely come when I move into my new house, and do a permanent setup in the garage, with proper ventilation there as well. I plan on running a couple of dedicated circuits directly from the panel for these 2, but I guess I need to figure out how much each pulls, so I know if I need one for each, or if one 12guage wire and a 20 amp breaker would do the trick.


Thermometer, I have been debating on what to do. I wanted to get one at least up and running, to get some throughput happening, but the other, I may go straight to TC4Shield to see how that works with it. I will be mounting one thermocouple in a hole drilled in the bottom of the pan, and another in the environment somewhere. Roast graphs and computer control sure would be nice...


I was very pleased with the second roast I pulled from it. I think I should be able to use the eyes, and the ol sniffer to get good roasts until I do something with temperature monitoring.

[Tex]

While sound is an excellent indicator of final roast temps*, I find a thermometer readout of the caramelization and 1st crack durations to be the very useful.

*For me, the 1st outliers of 2nd crack indicates a FC+ roast level, and 10 seconds into 2nd crack is FC++.

[Everett]

Great information Tex, good to keep in mind. I have been doing research on thermocouples and probes and whatnot. Reading the long main SC/TO thread has helped.


I agree it would be better to get probes in there sooner rather than later. I also need to do more reading about how the caramelization stage, drying stage, etc, all affect the roast, as well as durations between points, etc.




Peter mentioned that he had tried all sorts of ways to fix thermocouples in there, and ended up with the nuts to weigh down the probe in the slot. I had an idea of drilling a hole in the bottom just small enough to put a standard soft thermocouple wire, and to RTV (Hi Temp, Food Safe) it in, from the underside. This way, its always in the beans, under the arms and can easily be removed and replaced if need be, and if the hole is left in there, beans won't fall through 


With the way both of mine dump, it would be easy enough to affix the wire so it does not get pulled out. Ideas? I suppose I still need to figure out an environment temp probe as well.

[peter]

I had an idea of drilling a hole in the bottom just small enough to put a standard soft thermocouple wire, and to RTV (Hi Temp, Food Safe) it in, from the underside. This way, its always in the beans, under the arms and can easily be removed and replaced if need be, and if the hole is left in there, beans won't fall through 


With the way both of mine dump, it would be easy enough to affix the wire so it does not get pulled out. Ideas? I suppose I still need to figure out an environment temp probe as well.

With the way you're going to dump, that should work fine.  I tried the hole/tube in the base, and threading a t/c through that; it got good readings, but I had to take it out each time to dump.

It might be hard to go under the arms, yet in the bean mass, w/o getting a reading from the metal bottom plate.

I'd forget about the ET probe until you know you need that.  It'll be the easiest to do, whenever.  In that regard, remember that the air circulating down the outside circumference will be warmer than the air going back up the middle to the TO, so where you put both probes will read differently from the outside to the inside.

[Everett]

This is true about getting a reading from the bottom plate. I hadn't thought about the heat retention in the teflon/metal. Perhaps I will do a solid probe through the bottom, and bend the arms so they just clear it. If I bend them too high though, then I may need to extend them somehow, so they still reach the edges. Right now, with the stock arms, it stirs 1lb super nicely. If I were to stick a thermocouple up from the bottom, i could have it flop over to the side...Just thinking out loud here.



I will be shooting to do 18-20oz, to have a solid 1lb after roasting, so it won't be often that I roast smaller batches, and if I do roast smaller batches, I think I would be less reliant on the temps as well. I've gone this long without em.

[mp]

Looking good Everett!

 

[smico]

Everett, it looks very very good.  Obviously you are man of action.

I highly recommend TC4.  You can extend it later and eventually build full control of the roster.  I enjoy this setup with my Hottop.  If you fix the thermocouple to the roaster, make sure to do thermal insulation of the thermocouple.  I used piece of silicon spatula stolen from my second drawer.

[Everett]

Thanks for the kind words Miroslav   To be honest, with some things, I am a man of action, such as coffee roasters, but with others, not so much. For instance, I am in the middle of a house renovation that has been going on for too long now, haha.


I'm curious what you mean my insulate the wire with silicone? I may have a different picture in my head than you. Do you mean to just protect the wire from the bare metal to keep it from getting cut?


I had not thought about going TC4 at the moment for just temperature control, I was thinking the whole 9 yards, with 2 Solid state relays controlling the heat from the TO, and one other thing. Agitation? TO Fan speed? I hear people saying that you dont need to adjust the to fan speed, but I cant help but think this may be beneficial in some way. Either way, I suppose I can add the SSRs in the future, and just use TC4 for temperature monitoring at the current time. Time to research some pc apps. I've played with RoastLogger, and Artisian. Any other recommendations?


Are you able to have completely automated repeatable roasts with your Hottop Miroslav? Say you create a template by doing a previous roast manually, can you load that template and click a button to have the Hottop do the same thing?

[Everett]

Double Post!


I was just doing some reading on Home-Barista about Artisian for anyone interested, and read an interesting post about thermouples vs thermocouple probes.


Arpi says: "It is best to use a thermocouple probe (needle with thermocouple inside), especially it you are interested in delta BT. A naked thermocouple is very sensitive and fluctuates rapidly with currents of air, creating jitter. Meanwhile a prope (long needle) has a shell that acts as an averaging filter."

Perhaps I should be looking at mounting a needle from the bottom of the pan. This may average out the temperature readings enough that even though it is in contact with the bottom of the pan it would still be accurate. That said, if the computer software only samples every X Seconds, it may create a nice curve anyways even if it was a naked thermocouple and it was all over the map.

[smico]

I'm curious what you mean my insulate the wire with silicone? I may have a different picture in my head than you. Do you mean to just protect the wire from the bare metal to keep it from getting cut?

I use metal shielded thermocouples, and originaly I attached them to the hot metal parts, which caused not very responsive readings.
I posted some photos of my installation in another forum:
http://www.homeroasters.org/php/forum/viewthread.php?thread_id=2785&rowstart=5

I had not thought about going TC4 at the moment for just temperature control, I was thinking the whole 9 yards, with 2 Solid state relays controlling the heat from the TO, and one other thing. Agitation? TO Fan speed? I hear people saying that you dont need to adjust the to fan speed, but I cant help but think this may be beneficial in some way. Either way, I suppose I can add the SSRs in the future, and just use TC4 for temperature monitoring at the current time. Time to research some pc apps. I've played with RoastLogger, and Artisian. Any other recommendations?

TC4 will be perfect for you.  You can buy a kit and solder loose components yourself.

There is still no perfect logger.
Artisan - Pros: Nice graphs, ROR, multiple temp curves on one graph  Cons: No support for HTC controller for full automation
RoastLogger - PROS: Support for HTC controller - full automation for me  Cons: Only ET and BT curves, no ROR
Roasterthing - PROS: Coffee Inventory  Cons: Does not support Arduino-TC4

Are you able to have completely automated repeatable roasts with your Hottop Miroslav? Say you create a template by doing a previous roast manually, can you load that template and click a button to have the Hottop do the same thing?

Yes.  I save every single roast, and then I can replay to get most of the times identical curve.  I use Jim's Hottop Controller (HTC), and Roastlogger.

[sea330]

Thanks for going to the trouble of posting the details of your project, very nice job. Paying it forward with great ideas is what GCBC is all about. I have a 2 Kilo commercial roaster that I use to feed friends,family,neighbors, up to about 70 lbs a month, sure makes it easy to deal with that amount with this roaster. I spend hours in the shop playing with my home builds. I have modified several SC/TOs for friends getting them started in roasting. We co-op green purchases getting larger quantities at better prices, been working great for us. Having the ability to purchase half and full bags of coffee makes it nice. Good luck Rich

[Everett]

Hmmm, I see what you mean about no perfect logger. Artisian is nice as far as I can tell, but the ability to replay roasts in RoastLogger sure is tempting. That said, I think learning to roast based on RoR would be super beneficial.


Thanks for the link to your build details. I suppose to do this accurately, I would need to secure the probe into a hole in the bottom of the SC pointing straight up, and have it stick out at least 3/4" no? The thermocouple sensor is usually in the tip of the probe correct? So it would not get readings from the hot plate it is mounted through if it were to stick out a certain distance? I wonder also about drilling the hole bigger, and securing it with hi-temp RTV, so the probe doesnt even touch the metal surrounding it. Then it would not need to stick out as far? Is this what you meant by insulating it? Having Silicone between the hole edges and the probe? (Please Excuse the awesome MSpaint Skills)

[smico]

You might be right.  I noticed major difference in response for ET, but BT was quite similar.  Still I get more precise temperature now.
And by insulating I mean that your probe touches material that has low temperature conductivity, whatever it is.
As you can see I glued big nut to the roaster (JB Weld), and screwed in piece of silicone, drilled a hole in silicone and pushed the probe through.

You can try both loggers.  I use them both as I have two TC4 sets, one for "production" and spare one for experiments and tests.  Not selling it, don't ask.