Who is here? 1 guest(s)
 Print Thread
Skywalker roaster mods
renatoa
Before 3D printing there was tinkering/tinsmithing...and before that there was... origami Grin
I prototype mine from cardboard using scissors, before transposing on a tin sheet.
 
ctjameson

Quote

renatoa wrote:

Before 3D printing there was tinkering/tinsmithing...and before that there was... origami Grin
I prototype mine from cardboard using scissors, before transposing on a tin sheet.


I should have done cardboard aided design, but I have a stack of old filament that will be basically useless soon if I don’t use it for something. Prototyping seems to be the best use case for it.

The funnel is like 99% done. I test fit the most recent full size model today and it’s reeeeaaaally nice. I think when I was trying scale it down to just 100x100mm, I shrunk the bucket a little too much. I’m going to bump it out a smidge for some breathing room, but the funnel design is ready otherwise.
Skywalkerino - Arduino/artisan powered Ali Roaster
 
mhsieh66

Quote

ctjameson wrote:

Final design of the funnel entry solidified. Ready for a final full print prototype, then print in food grade PETG. Will post the final STL tomorrow if the print goes well tonight. It’s only 12 hours of printing. 🙃


Good job! Question, it looks like it needs to use support for one of the face from the loft 1, its angle is 57º, do you consider creating another version that doesn't need to use support?
 
ctjameson

Quote

mhsieh66 wrote:

Quote

ctjameson wrote:

Final design of the funnel entry solidified. Ready for a final full print prototype, then print in food grade PETG. Will post the final STL tomorrow if the print goes well tonight. It’s only 12 hours of printing. 🙃


Good job! Question, it looks like it needs to use support for one of the face from the loft 1, its angle is 57º, do you consider creating another version that doesn't need to use support?


Zero supports required on my jank 3 year old Ender 3 with the upgraded hot end. Print honestly comes out far nicer than it has any business being without any supports.
Skywalkerino - Arduino/artisan powered Ali Roaster
 
ctjameson
So I took the original setup and cleaned it up a lot. Designed a case for the header-less model Leonardo and made it a USB-C female instead of a cord attached directly to it. This makes it much more modular. I’m super happy with how it turned out. I’ll release the files later this afternoon when I get back to my computer. Sweet Maria’s delivery today, going to do a couple lbs tomorrow and report back.
ctjameson attached the following images:
img_3215_large.jpeg img_3212_large.jpeg img_3211_large.jpeg img_3210_large.jpeg img_3209_large.jpeg

Edited by ctjameson on 01/13/2024 5:55 PM
Skywalkerino - Arduino/artisan powered Ali Roaster
 
Robotic Kitten

Quote

ctjameson wrote:
and made it a USB-C female instead of a cord attached directly to it.


I do like this idea. I had a bad usb cable, which I cannibalized for this project, and incidentally I've also ordered the same USB-C connectors. Going to redo with USB-C. I'm currently using Arduino Nano. Tested it and it works, but haven't done a real roast yet. BTW, is there anything I could use to simulate coffee beans? Or should I just buy the cheapest beans for experimentation?
 
ctjameson

Quote

Robotic Kitten wrote:

Quote

ctjameson wrote:
and made it a USB-C female instead of a cord attached directly to it.


I do like this idea. I had a bad usb cable, which I cannibalized for this project, and incidentally I've also ordered the same USB-C connectors. Going to redo with USB-C. I'm currently using Arduino Nano. Tested it and it works, but haven't done a real roast yet. BTW, is there anything I could use to simulate coffee beans? Or should I just buy the cheapest beans for experimentation?


Do you mean to “practice” roasting? I’m not familiar with anything like that. I think it’s just trial by fire with good product.
Skywalkerino - Arduino/artisan powered Ali Roaster
 
mhsieh66
I modified ctjameson's funnel design to a small one but still be able to fit 500g green coffee beans, and it's safe to print it without any printing support.

https://cad.onsha...4d0c98fc47
 
ctjameson

Quote

mhsieh66 wrote:

I modified ctjameson's funnel design to a small one but still be able to fit 500g green coffee beans, and it's safe to print it without any printing support.

https://cad.onsha...4d0c98fc47


Yeah I tried to keep it well within the 150x150x150mm for smaller printers. I confirmed today that the most recent revision can for sure hold 500g of green, but I went ahead and just added a few mil vertical for some "breathing room."

File here:

https://thangs.co...lView=true
ctjameson attached the following image:
img_3221_large.jpeg

Edited by renatoa on 01/14/2024 2:42 PM
 
renatoa

Quote

ctjameson wrote:


Do you mean to “practice” roasting? I’m not familiar with anything like that. I think it’s just trial by fire with good product.


chickpeas.
Here they are 5 times cheaper than greens and equivalent thermal, missing first crack and development phase though.

Another alternative is using beans with defects, resulting from a pro roastery sorting process, if anyone known in reasonable distance.
My first three roasts on Skywalker were using such beans.
 
Robotic Kitten

Quote

renatoa wrote:
Another alternative is using beans with defects, resulting from a pro roastery sorting process, if anyone known in reasonable distance.
My first three roasts on Skywalker were using such beans.


Seasoning beans. Coffee coral offers for about $3/lbs + $12 shipping. I'm considering it.

Quote

Do you mean to “practice” roasting? I’m not familiar with anything like that


I'm planning to mess with the PID and direct control of the roaster, similar to TC4 and need a bean mass, especially with direct IR roaster like Skywalker
 
renatoa
How many successful PID roaster control cases do you know first hand ?
Especially direct IR...
 
Robotic Kitten

Quote

renatoa wrote:

How many successful PID roaster control cases do you know first hand ?
Especially direct IR...


Successful is too broad of a definition. Kaleido follows the temperature profile quite closely, however it does not mean it would follow the "burner" or RoR profile closely. Anyway, this is just an experimentation and a playground to get my hands dirty with PIDs

BTW, based on your "Celsius Patch" from https://homeroast...post_77899 I've created a pull request https://github.co...ter/pull/2. Any volunteers to test it?
 
ctjameson

Quote

Robotic Kitten wrote:

Quote

renatoa wrote:

How many successful PID roaster control cases do you know first hand ?
Especially direct IR...


Successful is too broad of a definition. Kaleido follows the temperature profile quite closely, however it does not mean it would follow the "burner" or RoR profile closely. Anyway, this is just an experimentation and a playground to get my hands dirty with PIDs

BTW, based on your "Celsius Patch" from https://homeroast...post_77899 I've created a pull request https://github.co...ter/pull/2. Any volunteers to test it?


Confirmed able to switch back and forth between C and F now. Will have to test actual temp data values later when I roast. Thanks for this update!
Skywalkerino - Arduino/artisan powered Ali Roaster
 
renatoa
You're welcome.
 
ctjameson

Quote

renatoa wrote:

You're welcome.


Confirmed temps are changing with correct deltas.
Skywalkerino - Arduino/artisan powered Ali Roaster
 
Mike_Mathis

Quote

ctjameson wrote:

Quote

mhsieh66 wrote:

I modified ctjameson's funnel design to a small one but still be able to fit 500g green coffee beans, and it's safe to print it without any printing support.

https://cad.onsha...4d0c98fc47


Yeah I tried to keep it well within the 150x150x150mm for smaller printers. I confirmed today that the most recent revision can for sure hold 500g of green, but I went ahead and just added a few mil vertical for some "breathing room."

File here:

https://thangs.com/designer/ctjameson/3d-model/992291?manualModelView=true


Are you going to post the file for download in this thread?
*Kaffelogic Nano 7
*Skywalker V1
*2023 BC-2
 
ctjameson

Quote

Mike_Mathis wrote:

Quote

ctjameson wrote:

Quote

mhsieh66 wrote:

I modified ctjameson's funnel design to a small one but still be able to fit 500g green coffee beans, and it's safe to print it without any printing support.

https://cad.onsha...4d0c98fc47


Yeah I tried to keep it well within the 150x150x150mm for smaller printers. I confirmed today that the most recent revision can for sure hold 500g of green, but I went ahead and just added a few mil vertical for some "breathing room."

File here:

https://thangs.com/designer/ctjameson/3d-model/992291?manualModelView=true


Are you going to post the file for download in this thread?


https://homeroast...post_78104
Edited by renatoa on 01/14/2024 2:41 PM
 
Mike_Mathis

Quote

ctjameson wrote:

Quote

Mike_Mathis wrote:

Quote

ctjameson wrote:

Quote

mhsieh66 wrote:

I modified ctjameson's funnel design to a small one but still be able to fit 500g green coffee beans, and it's safe to print it without any printing support.

https://cad.onsha...4d0c98fc47


Yeah I tried to keep it well within the 150x150x150mm for smaller printers. I confirmed today that the most recent revision can for sure hold 500g of green, but I went ahead and just added a few mil vertical for some "breathing room."

File here:

https://thangs.com/designer/ctjameson/3d-model/992291?manualModelView=true


Are you going to post the file for download in this thread?


https://homeroast...post_78104


Thanks anyway. I prefer to not have to create an account to download a file.
*Kaffelogic Nano 7
*Skywalker V1
*2023 BC-2
 
ctjameson

Quote

Mike_Mathis wrote:

Quote

ctjameson wrote:

Quote

Mike_Mathis wrote:

Quote

ctjameson wrote:

Quote

mhsieh66 wrote:

I modified ctjameson's funnel design to a small one but still be able to fit 500g green coffee beans, and it's safe to print it without any printing support.

https://cad.onsha...4d0c98fc47


Yeah I tried to keep it well within the 150x150x150mm for smaller printers. I confirmed today that the most recent revision can for sure hold 500g of green, but I went ahead and just added a few mil vertical for some "breathing room."

File here:

https://thangs.com/designer/ctjameson/3d-model/992291?manualModelView=true


Are you going to post the file for download in this thread?


https://homeroast...post_78104


Thanks anyway. I prefer to not have to create an account to download a file.


It's SSO with like 3 platforms you probably already have an account for, but your loss ¯\_(ツ)_/¯

You can also open the onshape link I posted and export it as an stl yourself if that's still not acceptable.

Thanks!
Skywalkerino - Arduino/artisan powered Ali Roaster
 
Mike_Mathis
This evening I added an alternate thermocouple to the roaster. Thanks Josh for your video on probe location. It helped tremendously with my decision of where to place it. I did a dry run test in manual mode with a Mastech connected to do a comparison between probes. My test concluded the factory probe has lag compared to the thermocouple with Mastech. As you can see from my pic, the probes are the same diameter and the same length into the roasting chamber. I found it a little interesting that both probes are obscured from a direct line to the heater except when the trier is out. During PH, the factory was a consistent 7-8° behind the Mastech reading. When it settled at 180 setpoint, the factory caught up and read pretty much the same. If the TC is more accurate, when the factory shows you are charging at 200, you may be closer to 210. Definitely a slow read from the factory probe.

i.imgur.com/YJCTbOS.jpg

i.imgur.com/N8kAGq0.jpg

i.imgur.com/dCCyeHA.jpg
*Kaffelogic Nano 7
*Skywalker V1
*2023 BC-2
 
Dan N
I've done yet more analysis on Josh's thermistor data in post #3 and have learned a great deal about the inner workings. I'm now able to calculate the coefficients for the Steinhart–Hart equation to match the data: a=7.119210e-04, b=1.984603e-04, c=1.054138e-07

I wouldn't use the equation directly on the Arduino except for testing. A lookup table would be faster and allow adjustments.

I believe the thermistor is either 220k or 230k based on the data and also my thermistor measured 303k at 66F (18.9C) which fits the curve for those.

The controller appears to handle temperatures as integers in Celsius units and truncates the fraction when displaying Fahrenheit: 51C=123.8F displayed as 123. That's why some temperatures appear to be missing in Josh's data: 11C=51.8F, 12C=53.6F (no 52F).

I also have a pretty good idea for how it makes and uses the 2 readings and when it switches between them.
 
ctjameson
I roasted 3 pounds tonight of a Nicaraguan from Sweet Maria's tonight. Roaster control worked flawlessly, however either the probe temps are low from the actual bean mass or the Celsius data being presented by the TC4 code isn't correct. My bean end temps were really low. The first roast, I stopped at 215 C and it was well into a darker roast by that point. My next two roasts ended at 185C and are right where I want on the roast scale of things.

Really enjoying playing with this little guy. I'm especially happy with the approachability of the machine. I have multiple friends considering getting into roasting now simply because of the affordability of this thing.
ctjameson attached the following images:
img_3245_large.jpeg nicaragua_third_roast.png img_3248_large.jpeg nicaragua_first_roast_2.png

Skywalkerino - Arduino/artisan powered Ali Roaster
 
renatoa
Please check my post about IR roasting and temperature range in the main Skywalker thread. FC around 180 is the norm here.
Manufacturer states in the manual 170-190 as FC range, and he is right.
Any attempt to compare this roaster ranges with the classic values found on the internet is a mistake, imo.
Trying to stretch the profile to match the other machines is a mistake even bigger.
Edited by renatoa on 01/15/2024 1:48 AM
 
renatoa

Quote

Dan N wrote:


I believe the thermistor is either 220k or 230k based on the data and also my thermistor measured 303k at 66F (18.9C) which fits the curve for those.



This is absolutely weird... mine is 5K !!! Shock
Connected with a 30k upper resistor in a divisor, and powered from 5V, this give about 0.7V voltage at ambient on NTC pins.
 
Jump to Forum: