You are moving along nicely! The three motors in a row are impressive. To me, it looks like your ceramic heating element holders are spaced too far apart. When the nichrome heats up it expands and becomes weak, so it slumps. If it touches the bottom plate it'll arc and burn out. Morelectric doesn't have the perfect ceramic holder. I made mine from their round bushings. I used stainless wire to connect two, one as a bushing for the heater, the other as a riser to create a standoff. You could insert a standoff between each of your corner holders. You might need two risers to match your heating element height.
Thanks Dan... maybe "Sensei Dan" is more appropriate. Your guidance is much appreciated.
I share the same concern about the droopy coil. I just ordered some ceramic rings. For now I I'll install a couple more standoffs between the longer runs. Which can be re-used with the rings once they arrive.
Next on the to-do list is the thermocouple probes. Originally I wanted to install one in a similar location to Dan's in the photo above. That location is great if the drums are all skinned the same. In my case the large drum is skinned in stainless mesh, and the small are nearly solid. For a given "E" temp - I think the temperature the beans experience will be significantly different when using the one drum vs the other. So two different locations for temperature readings are needed. The large drum can have a thermoprobe in the 3 or 9 o-clock position and set 1/2" away from it. The small ones should have a probe inside the drum near the top.
Here's the bent 1/4" OD aluminium thermoprobe for the small drum:
That should give a better reading of the "E" temp the beans are actually exposed to without being too close to the bean mass (relative). If it fails miserably at giving a good "E" reading it can be used as a bean temp probe instead.
I hope this doesn't cause too many problems for the PID controller. When using the large drum the response time will be faster than small ones. Hopefully not too much of a difference or there will be overshoot undershoot problems. Setting the damping factor somewhere in the middle should work. With the user (me) making adjustments to compensate.
Wonderful! Can't wait to hear if it actually works. :) You've solved the "melt the motor cable on the heater box while dumping" problem using those spring strain reliefs. One end of those coiled cables disconnects to permit switching to the large drum, right?
What I did on mine was install a stationary plug and socket for the motor power. When I tilt the drum power is disconnected. No cable at all.
That's not romex on the heater box is it!? While we can do whatever we want on homebuilt machines, romex is not rated for unprotected applications. I'd put in 1/2" conduit (you have the right size holes already) and stuff it with high temperature wire from MorElectric.
Just looking at this great machine with the fine craftsmanship has me thinking. We need to have a new roaster of the year competition. Great machine, now I want too see smoke coming out and some of the shine turn dark, think it will roast great.
All you need in life is ignorance and confidence, and then success is sure. Mark Twain
Gotcha Dan. I aim to please in both aesthetics and function.
Wired up all the innards. Then ran this little test:
Warning: Very Boring Video You will lose 10,000 brain cells guaranteed. YouTube Video
The test was a ramp from 300f to 450f. In reality 420f should be the stopping point.
That would be around the 4 min mark. I call that a success considering the only thing that's insulated/sealed/double-walled is the heater box. The rest has lots of air gaps.
Dan, I'd like nothing more than to shake your hand. Thanks.
I'm surprised by the efficiency of it. First crack in 10.5 min. is ideal.
Looking at the front with its large openings I was a bit skeptical.
It appeared too lossy for electric heat, had propane burners in mind as backup.
Luckily; I'm self employed, dedicating large blocks of build time was no problem.
Dangit! You figured out why cooling was taking 2min instead of 1:57. I was wasting 3 seconds doing things in the wrong order!
Metal Goods 80
Wiring,Connectors Etc. 30
Controller, Ssr Free
Labor 1 (100hrs @ $0.01 an hr.)
Total (Me) 211
Total (Customer) Crazy Expensive
I'm as surprised as you with how well it works from the get go. It wasn't without many revisions along the way. Some part ended up a 16th out of square - so I had to make a new one. The little drums required more assembly/disassembly than anything else. Shaving a little off here, re- drilling/tapping there, on and on.
I hear the cracks loud and clear. I think the white noise from the motors overwhelms the microphone on my cheap camera making them muffled.
I spoke too soon. There is a flaw with the design. When using all three drums at once the beans in the middle drum roast significantly faster. Using the same beans in all the drums as a control, 1st crack in the middle drum starts 2 min before the others. Which makes profiling and temperature drop to extend the time between cracks impossible. They are too far out of sync.
I'm wondering if strategically positioned baffles under the middle drum will help with this. Can't think of any other way at the moment. Any Ideas?
Yeh, I'd look at baffling first. If that doesn't work, then perhaps a convection fan. Worst case scenario you reconfigure the heater to be three identical units, but operating them simultaneously with your PID.
I see a three things in your design that would effect the heat dispersion. The heating element doesn't have a uniform pattern across the three drums; the inside double wall in the coil section reduces the exposure of the two outside drums to the hot air around the coil; the middle drum has two drums insulating it from the outside air - or stated the other way - the outside drums have more exposure to cool air. Without hearing your experience, I would have expected these to offset somewhat.
Reflecting the heat away from the middle drum or insulating the shell around the outside drums to reduce their heat loss, especially the upper left and right corners by the two outside drums - may help. If this doesn't work - short of separate controls, you could also scale back to two drums with symmetrical layout. Getting mechanical symmetry with three drums may be difficult.
All that being said, I really admire your metal working skills. I've never done sheet metal work, and I'm inspired by what you've done.
Great project! What do you think the budget worked out to be? (if I may be so bold) -Scott
So far - my guess is around 250.00 Which will jump to 350.00 once the large-drum "top" is built.
Thanks for your feedback Dan and John,
It's funny, when I first explained the uneven heat distribution problem to my girlfriend, her immediate response was "I could've told you that". Normally, she's the opposite of an engineer-type thinker.
I managed to reduce the lag between drums by installing baffling. It's an ongoing process to find the right configuration. At first, a solid sheet was placed under the middle drum, that led to the outer drums finishing in sync, but well before the middle. Next, strips of aluminium were placed under the middle drum creating a 50% coverage. Strangely, that caused the middle and right drums to finish much faster than the left. So, a strip of steel was placed perpendicularly atop of the aluminium strips on the right side. That brought things much closer to the ideal balance. I'll do a few more roasts tweaking the location before welding on a permanent baffle.
We finally have parity between all three drums. It required modifying the drums and the heating element layout:
Stainless #6 mesh with 1/8" gaps now lines the drums, allowing more air pass-through. Which also makes them much more responsive to temperature changes.
The element was moved away from the center to counteract the tendency for the heat accumulate there, leaving the outer edges colder.
That did it. Now we can roast profiles. Yay!
Next, work has begun on the 2nd "top" for doing larger batches. Call me crazy, but I decided to split the large drum in two. The single drum "top" was both ugly and heavy:
Here is the double barrel shell and drum:
Much better. Now it will tip to the front just like the 3 barrel version. The drums will get the same stirring vanes and mesh lining as the small drums. That should make the performance (RoR) similar between the different "tops".
That brings the total number of drums up to 5, how did this happen?