FWIW, I did two 6 oz roasts the other day and it didn’t get stuck. First roast was an attempted repeat of my initial attempt. (I guess I pulled them too soon, not as dark as the first attempt, I roasted to the same time thinking they’d match. I guess just go by color next time). The 2nd batch had tiny beans, so less likely to get stuck anyway.
That is good to know that you are having better results. My wobble disc is in York PA now, so it should be here next week.
I have the mechanical and structural design done for the Rotisserie roaster and actually looks pretty good to me. I will put all the parts I have on hand together before deciding whether to order the rotisserie. Actually, this design if it works is in a lot of ways simpler than the flour roaster and should be less expensive. The only extra items I should have to buy is the rotisserie and a SS shaft.
Just looking at this again, I need to swap the motor and idler as the heat gun handle will hit that motor. That is an easy swap. Putting the assembly together does help to see some problems before actual assembly. I did not have 4 - V-slot 2020s, so used a couple of 2040s.
When looking closer at these cages they have square holes in them and not sure what size they actually are. Part of me just had to get this design out of my head. I still might play with the movement of these and maybe make a 3d printed plastic cage to actually test the idea.
Here is my crazy idea in motion. Should I do a destruction test to see how well the PLA rotisserie handles the heat? I would not try that on the deck though, maybe in the gravel driveway.
I have heard a rule of thumb that coffee should be roasted within 15 months. Ground within 15 days and brewed within 15 minutes. But those are obviously gross rules of thumb and stuff like the commercial bags and gas they use for beans on the grocery store shelf make a big difference.
My guess is there is no difference between 0-15 months and a difference that wxperts can sense above that. But that’s based only on my opinion of coffee people ![]()
I prefer the energy drinks
Coffee always upsets ![]()
I received my 316L SS 3D printed wobble disc today and looks like they did a good job with it. If I did it again, I would probably try to cut down the weight a little more. It is solid at 201 grams (7 oz). The 5.5mm hole I designed in for the M5 shaft is a perfect fit without having to do anything to the part. I did a test roast with it and did have an alignment problem with the motor and the shaft. The previous version using the crank handle was more forgiving, but I measured 2.5mm off in the Z axis. I added slotted holes for the motor this time just in case. I am printing a new mount and will have to reprint a new planetary mount as well unless I decide to drive the shaft directly with the motor. I can’t really add a slotted adjustment on the gear system, so will make sure this lines up first before correcting that. I connected all the hardware on the outside of the sifter 1st to make sure it would work and it does hold solid.
Think I am finished with this version of the Planetary Bean Bouncer. I had a couple of electrical problems. The 1st was a CNC shield was flakey on a motor connection somewhere. Probably bad solder joint somewhere. I had a couple of spares, and they are only $2.50 each so I just replaced it and that problem went away. Next time I am soldering I might just hit all those solder joints with a soldering iron rather than throwing it away. The 2nd electrical problem was one of the switch wires broke at the board. I was using ribbon cable for those wires which is probably 24 or 28 gauge. I redid the 4 switches with 22-gauge flexible silicon cable. That stuff bends really nicely inside this case. I did not redo the other LED or Buzzer wire since they are a little more trouble and not critical.
I also had one mechanical problem on the 1st roast on Thursday. I was originally using 1 locknut on the end of the M5 threaded rod connected to a plastic part on the top of the planetary gear. That nut came loose about 1 minute into that roast. To salvage the roast, I took out the beans that were burnt and added some to replace them and put 2 locknuts on there as jam nuts and got through 3 roasts that day. Since the brass coupling I was using between the motor and the sun gear has been working well, I used the same thing on a newly printed gear set. Here is the old style. I am also tying the gear to the motor mount together and adding route for the thermocouple.
Here is the coupling I am using. Amazon.com: uxcell Hex Coupler 5mm Bore Motor Hex Brass Shaft Coupling Connector for Car Wheels Tires Shaft Motor 4pcs : Industrial & Scientific
To give the set screws a better grip, I did decide to flatten the one end of the M5 shaft where both set screws engage. That filed down ok, and I can still thread all the hardware in from the other side.
Since I am using a bearing supported like a field goal on the back of the sifter, I decided to drill out those shaft holes in the flour sifter some so they will ride on the bearing and not hit the sifter wall. My neighbor had some cobalt bits and used a 3/8” bit. This extra space makes it easier to also connect most of the hardware on the shaft before putting it in the sifter. I made a drill template for inside the sifter to keep the metal shards from going in the sifter and more importantly made the sifter rigid for the drilling. That worked really well.
I changed the internal hardware some. Adding all those spacers was just too much trouble as I did not have much room to tighten the nuts at the end. I have 3 spare threaded rods, in case it is ever a problem. Starting from the bearing, I have a locking collar on the outside, serrated flange nut, hex nut, 18mm long spacer, serrated flange nut, hex nut, star washer disc. Same hardware on other side except for the locking collar is not needed there. I did need one really thin wrench to use with the flange nuts as that flange gets in the way with a wrench on the inside hex nut.
I also redid the frame using V-Slot 2040 on the bottom since I found some from another project to repurpose for this one. That makes the frame a lot simpler design. I also added handles mid-way into the center supports for grabbing when pouring the beans out. The center supports are not really needed for structure but do give a good place to grab on one side and give some mounting spots for the MAX6675 and cable ties for motor wire.
From my testing of the roaster with and without the planetary gear, it might not really give much advantage. I am a little partial to it, and it does work. I would also have to change the name of the roaster if I did away with it.
The firmware is working really well. I have 4 buttons with a short and long press feature on each.
#define BTN_MOTOR 10 // CW/CCW toggle + long-press Buzzer Mute (ON/OFF)
#define BTN_PROFILE 11 // Change to different RPM range + long-press (LCD reset)
#define BTN_TIMER 12 // start/stop + long-press reset timer to 0
#define BTN_RPM 13 // Change to different RPM + long-press to toggle F/C
I also just added a feature where it will give a warning double beep at 250F and triple beep at 275F. If I move that thermocouple further down, I will have to increase those values.
The display on the LCD gets garbage characters every once in a while. The LCD reset feature will reset the display without having to unplug/plug back in the Arduino. It might be getting some interference from the motor wires which run right beside them. I might could fix most of that problem with changing those 4 LCD wires to a shielded cable.
One of the problems with using aluminum extrusion with this coffee roaster is the chaff likes to get caught in the grooves. I’m using this Flexible Rubber Extrusion Strips (for 20mm Aluminum Extrusion) - Black, Blue, or White - ZYLtech Engineering, LLC that should eliminate most of that. I have had this for a few years and glad I never threw it away. I kept it away from where the flour sifter is actually making the extrusion hot.
To cushion the bottom, I added these felt feet.
Hey this is the one that pulled me out of the funk/depression after my mom died.
I finally have the initial release of these files published on github. I still need to do some work on the docs.
By some magic mine haven’t given up their grip yet. I await a ruined roast.
A leaf blower is my favorite chaff blaster.
I was commiserating with grok on adding artisan software control this morning.










