Who is here? 1 guest(s)
 Print Thread
6kW Electric Air Roaster
greencardigan
Hi all,

I have been progressing with my latest air roaster build and have reached a point where I'm getting close to trying some bean circulation tests.

My goals with this roaster were:
- Roast batch size of 1kg (2.2lb) roasted. 1.25kg green.
- Cooling in roast chamber
- Reasonably compact and portable.
- Cheap and easy to build with my limited tools.

To achieve these goals I'm attempting to use
- 3 x 1800W heat gun elements
- 600W vacuum blower
- Spouting fluid bed design
- 6 inch diameter roast chamber

Current state of the build.
forum.homeroasters.org/forum/attachments/img_20180224_175526.jpg

Electrical diagram
forum.homeroasters.org/forum/attachments/screenshot_20180211-231810_01_1.jpg
greencardigan attached the following images:
screenshot_20180211-231810_01_1.jpg img_20180224_175526.jpg
 
greencardigan
And some photos of the internals.

3D printed manifold
forum.homeroasters.org/forum/attachments/img_20180127_085512.jpg

forum.homeroasters.org/forum/attachments/img_20180127_174508.jpg

Vacuum blower
forum.homeroasters.org/forum/attachments/img_20180124_205838.jpg

forum.homeroasters.org/forum/attachments/img_20180222_193610.jpg
greencardigan attached the following images:
img_20180127_174508.jpg img_20180127_085512.jpg img_20180222_193610.jpg img_20180124_205838.jpg
 
allenb
Hi Brad,

Another extremely cool and nice looking roaster you're building. Nice re purposing as well!

The fan and heating elements assembly remind me of some rocket propulsion internals I got to see while at one of my previous places of work.

What's the three way splitter air distribution adapter made of?

I didn't read carefully to see 3-D printer so disregard question.

Allen
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
greencardigan
Hi Allen,
The air splitter is 3D printed in ABS.
 
sethyrish
Very nice. Love the manifold. It'll be great when you can print in ceramic at home!
I didn't know a zero cross is needed for the elements. I thought just for motor.
 
greencardigan
A zero cross SSR is a normal SSR. They switch automatically when the AC crosses zero to allow a minimum of one half cycle to pass through.

The motor uses a random fire SSR which will switch whenever it is given the signal. Thus it requires the separate zero cross detector so the microcontroller can switch the power a specified time after the zero cross. This allows part of each half wave to pass through.
 
sethyrish
Thanks for replying. I've been searching arduino motor control info trying to work it all out.
 
greencardigan
I've had some level of success with my first bean circulation test. The little 600W blower can lift 1.25kg with cold air (no heaters on).

I'm using a sink strainer for a perf plate which had 62 x 2.5mm holes. Circulation with this was marginal. I had blower on 100% and still had to help it start circulating.

Drilling most of the holes out to 3mm helped a bit. That should be about 40% increase in the open area. 100% blower now gets circulation start fairly reliably. And it will maintain circulation down to 80% power.

Roast chamber diameter = approximately 145mm
Perforated area diameter = 45mm = approx 10% of total area.
Open area = 26% of perforated area and 2.5% of total area.

Overall I'm still not real happy with the circulation. I think it needs to be circulating faster? And maybe the funnel is too flat near the bottom? I have a standard conical shaped funnel in the mail so may switch to that when it comes.

Any thoughts on perforated areas? Funnel shape?


Edited by greencardigan on 02/28/2018 12:53 AM
 
Will2
I think the circulation is slow.
It may be related to the slope of the funnel.
Because when I tested the circulating on my 45-degree funnel,
there was no circulation with the 700W engine.
I need a stronger 1200W motor.
What is the inside diameter of the glass cylinder in mm?
(My cylinder is 144 mm inside, batch size 1290g, the heater will be different).

up.picr.de/30935071wr.jpg
Viliam
 
Will2
duplication, deleted.
Viliam
 
allenb

Quote

Overall I'm still not real happy with the circulation. I think it needs to be circulating faster? And maybe the funnel is too flat near the bottom? I have a standard conical shaped funnel in the mail so may switch to that when it comes.

Any thoughts on perforated areas? Funnel shape
?


Yes, a little sluggish running ambient air temp but will be fine once heated. I prefer a more gentle spouting action over the bean cannon approach. I would definitely try out what you have now with a heated air stream prior to jumping to the alternate RC bottom. Even if it takes a couple of seconds to get the air hot enough to get flow started, shouldn't cause any problems from my experience.

Looking good!

Allen
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
greencardigan
Thanks Viliam and Allen,

I'll try with some heat when I get a chance. I may also try a few other perf plate options.
 
BenKeith
Yep, When mine move that slow, they tend to tip and scorch badly
That roaster looks more like something getting ready to go in orbit than something to roast coffee beans.
 
Will2

Quote

BenKeith wrote:
... they tend to tip and scorch badly...


That's exactly what I meant.
My friends from the Netherlands definitely forgive me, that I'm publishing a photo without their prior consent:

up.picr.de/31962945wd.jpg

Frans writes: Old test beans, very uneven in color.
I think the reason is that beans is coming back too long, and they are cooled on a cooler surface.
Viliam
 
fransgo

Quote

Will2 wrote:

Quote

BenKeith wrote:
... they tend to tip and scorch badly...


That's exactly what I meant.
My friends from the Netherlands definitely forgive me, that I'm publishing a photo without their prior consent:

up.picr.de/31962945wd.jpg

Frans writes: Old test beans, very uneven in color.
I think the reason is that beans is coming back too long, and they are cooled on a cooler surface.


I was unable to ever log into this forum, always being sent back to a login screen but now I created a new ID to participate.

This my message, expanded:

Quote

You need to know (it was mentioned in one of the videos on screen) that we were using up a batch of Indonesia Sumatra beans for testing.

They are very odd and one should actually first spend an hour to sort out the bad beans and after roasting another 20 minutes of sorting ? that is, with a 1.2kg batch as I usually had with these beans.

I published a blog about all the work involved and the superior result if one wants to take the time and trouble.

In the end we still had a batch of these beans and decided to just use these for testing. Not sorting anything.

So we charged all beans, unsorted and so they also came out multicoloured.


It is also not wise to try to analyse beans from photos. One has no 'grey card calibration' at hand so the camera 'corrects' the white balance, lighting can be warm/cold changing the view et cetera.

On ones own screen there is no way to conclude anything about such details.

If beans are very blackish and oily, yes then it is likely the roast lasted a bit long OR was a bit hot in the end. But even then I think it's best not to be the first to come forward with conclusions.
 
Will2
I apologize for an improper closure (I did not mean the multicolored beans, but on Tipping: a Scorching).
At the request of Frans, the following explanation is given:

Publishing the photo is fine
but adding a WRONG conclusion from a photo is not okay.
Edited by Will2 on 02/28/2018 6:06 AM
Viliam
 
greencardigan
Hi Frans, glad you made it onto the forum.

I?ve been thinking about my perforated area and I think it needs to be bigger. Others have suggested it should be 20 to 25% of the roast chamber area. Mine is currently only 10% and the resulting plume is thin.
 
renatoa
You can increase holes to 4mm without worry.
Or. better, replace with a mesh made using expanded tin method, the biggest airflow %.

www.metalsupermarkets.com/wp-content/uploads/2016/05/Fotolia_88356122_XS.jpg
 
allenb

Quote

greencardigan wrote:

Thanks Viliam and Allen,

I'll try with some heat when I get a chance. I may also try a few other perf plate options.


Brad,

I'm sure you've already scoured Dan's "Dan Bolllinger's Fluid Bed Tests" in the downloads section of our site?

if not, it has a lot of useful data on tests of various sized roast chambers and perf plates that may be helpful.

Allen
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
greencardigan
Thanks Allen, yes I've been looking at that. It's been helpful but I'm not fully understanding it yet.
 
fransgo
Made a new video this morning:

https://www.youtu...DzR6C7zoV0

 
specialreserve
very nice. How does your system work it appears to be rectangular? I like the bean dump feature too.
 
fransgo
Thanks!

Yes, it's a roast chamber between these vertical plates, with a curved wall in between the plates to guide the bean flow. Tije recently narrowed the top bit inside to there is an easy airflow but no beans flying out with the chaff.

Some details of the setup are on my blog kostverlorenvaart... and on Jan van der Weel's blog www.byleew.nl
Edited by allenb on 03/11/2018 1:26 PM
 
greencardigan
A little update.

I receive an alternative funnel for the base of the roast chamber and it works slightly better than the previous one. I can get 1.25kg circulating at 60% blower power compared to 70% with the previous base. This was with the 61 x 4mm holes. I hope to try a larger perforated area soon to get a thicker plume.
Edited by greencardigan on 03/27/2018 1:20 AM
 
greencardigan
Another circulation test with the steeper funnel and a different perf plate (50 holes with 310mm2 open area).

I think the circulation is improved but minimum blower power required to maintain circulation was back up to 80%. But with the more restricted perf plate, the airflow is probably similar to the more open plate at 60% blower?



greencardigan attached the following image:
img_20180330_151731.jpg
 
Jump to Forum: