Hello,

Some time ago youtube suggested me a movie:
I was surprised how easy it such a roaster can be build, so I started to watch everything that I can find about the topic, ie http://www.cyberelectronics.org/?p=699

At this point i knew that i will make heat gun based fluid bed controlled by Artisan. Most information about it I found in here:

http://www.sinobi.dk/henrik/coffeeroaster1/

I bought second hand heat gun, disassemble it and take heater with pipe out.



Next was a mattress pump. It needs to have enough power to lift the beans, and be DC to be able to control it by PWM Mosfet (i dont know what would be the best way to controll AC blower). I found Intex 66626 - which has reasonable price , 78W of power, 600 L/MIN air flow (but don't believe in that) and have 4 different sockts to connect it to our heater pipe.

https://www.amazon.de/Intex-66626-Quick ... B000HBMFU0

Borosilicate tube 80mm 15cm - that you need to find locally, I just typed "laboratory glass [City name]" and try to find small workshop which will cut it for you. Another way is to find something like this:

https://sklep-chemland.pl/en/zlewka-wys ... mland.html

and cut it with diamond disc - but then there is a big risk that it will break.

Last thing to connect it all together is Inox Shaker, most of the 500ml are 80mm.



I've ended up with something like this:
Let me know if you want the whole story, how to connect it all together, parts prices, electronics etc.

So now I have it all connected to the Artisan via tcp modbus, with esp8266 module (I can share the code i wrote if anyone is interested). But what next?

What do you suggest to change in Artisan configuration for that kind of machine? Should i get a first profile from which i should start adjusting it to my needs, or just use sliders and then change it later based on that. I have never roasted coffee with any software, just very small batches in popcorn machine, but they were really bad

I'm planning to adjust pid settings based on this: https://github.com/marcov/pidtune

I see that no one is interested, but I learn a lot here, and I would like to give something back.

https://bitbucket.org/dduniec/coffeeroaster/src - ESP8266 firmware, you need to use PlatformIO to upload it.

Connection diagram and Artisan configuration can be found in attachments.

Blower test:
First roast (now i've added secont thermocouple on top and connected it to Artisan):

Air roasters produce different curves than drum roasters.

Search the threads here for the SR540 and SR800 and you will find many.

They are more likely to offer you advice on how to roast this way.

If you want to learn about Artisan settings once your devices are set up, there is information in the Quick Start Guide linked in my signature.

If you have questions about controlling your roaster I would search the Artisan Blog first then if not post in the Discussions section of GitHub (new feature of GitHub).

Thanks for contributing.

Michael

I use a fluid bed for a few years before switching to a drum roaster. Ideally, you want a probe immersed in the bean mass, yet away from the incoming airflow if you want to even attempt to compare roast profiles with other roasters.

Why did you switch? Drum machine is better?

I have one probe a bit on the side on bottom where the beans are, second on the top, with case reduced on grinder (quicker but less stabke results)

Do you think 400g is too much?

Dave .... very cool. Good work and thanks for the info.
I would say that 400g was too much. A big concern with the fluid bed roaster is having uniform bean movement (fluid). The pulsating that you see will make some beans in contact with the hottest air more than others and result in uneven roasting.

In fact, I've seen threads where people go to extreme lengths looking for optimal roaster chamber angles and shapes to get the fluid dynamics dialed in. The Australian forum, Home Roasters, seems to have a lot more of this style roaster.

I cant speak to Alan's motivations for switching from fluid bed to drum, but I think capacity is limited on the fluid. It is easier to buy a commercially produced drum roaster with capacities between 300g-50Kg. I dont know of many mainstream fluidbeds that sell in that space.

There is also a difference in flavor profiles. The two technologies have their own advantages and disadvantages. The biggest problem with fluid beds is that heat and bean movement are very strongly connected. Drums separate these two factors. The biggest advantage is that fluid beds can increase/decrease heat very very quickly. Drums are like steering a large ocean ship. Fluid beds are like a small bird flying.

I keep saying that I want to build one. I started with a heatgun breadmachine and keep wondering if some of those amazing first roasts were because of the roaster ... or it was just the magic of a new hobby. I would love to have something that could rip through 4 sample roasts on a pound of coffee so that I could test more before buying large quantities of my favorite beans.

They are way more straight forward to automate and are more portable too.

Good fun, in general. Thanks for sharing!

Rickpatbrown thank you for your comment i think i have a problem with this movement with my new machine

I wanted to have bigger and nicer machine, so i bought bigger tube, pt100 sensors and build new machine. Blower can easily handle 2kg even on not full power, heater (from heat gun) is rewired and have now 4300W. But I can't reach profiles from my 250g machine. Im running test roasts with 1kg, it's strange because it can't follow profile from the little one, but at the end drying phase and fc are way more quicker than 250g. Should i stop looking at previous machine and just start to develop new ones?

First run (3600W heater):



First roast:



Second run (4300W):




Second roast:



1kg bean movement:

http://www.youtube.com/watch?v=7-E9mJnRAVg