Not a member yet?
Click here to register.

Forgotten your password?
Request a new one here.
You must login to post a message.

06/04/2023 7:48 PM
alicb7 Welcome

06/04/2023 4:02 PM
Mooney Welcome

06/03/2023 10:25 AM
CJbeans Welcome

05/31/2023 12:14 PM
Welcome, JavaJoeJaybird

05/28/2023 6:19 PM
Laurierejoice Welcome

In Memory Of Ginny

Latest Donations
dmccallum - 10.00
JackH - 25.00
snwcmpr - 10.00
Anonymous - 2.00
Anonymous - 5.00
Users Online
Guests Online: 3

Members Online: 0

Total Members: 7,872
Newest Member: alicb7

View Thread

Who is here? 1 guest(s)
 Print Thread
Another electric 500g home roaster
I present here my cheap and ugly roaster. I?m ashamed, but I?m no craftsman and my workroom is very limited. My first concept was too complicated, 500 g electric drum roaster with additional preheating of incoming air. I reworked it, the low half of the roaster was cutted off with a grinder and made new.
And I welded for the first time, the 1mm metal plate of cyclone was a learning project for me.
So it looks as it looks. I am from Europe and my English is not top too.
So, the roaster is now fully functional and works very well. First advice : keep it simple!

My idea ? my PIDed Gene Cafe was good, but the attempts to make the internal probe were not fully satisfying. So I decided for a true drum roaster:
300-500g Capacity , good chaff removal and cooling behavior . No ?hobby home drum roaster? .
In other words : ?professional? roaster with chaff cyclone and home price. (own work not included)
No filters, all in one solution, ?plug and play?.

All stainless, 1600W heating, 180x180mm perforated drum, 1.5mm thick, holes 4mm in diameter.
Front and back sides are 3mm thick, double shell 1.5mm, mineral insulation 20mm.
Front side and perforated plate for drum are bought laser shaped. The rest is handy work.

I hope a perforated drum is better for electrical heating, it can be more responsive for heat changes.
I decided to have separate venting and loading pipes, it is simpler.

I?ve a bean probe , ET and MET (switchable). Now I use only BT and MET.
I?ve made for Gene ramping PID SW, I wanted to use it again. But PID ramping has here no sense. (compared to Gene , the air temp has too slow reaction on heating) Instead of it, I use only regulated output by PWM (and 40A SSR relay) , monitoring of temperatures and adjusting airflow. The SW shows in 15 sec intervals the minute increase of BT (on the first row left).

The motor from is 1/30Hp, 60RPM. It is only for 5 minutes intermittent use. I use a fan to cool down it. In 30?C air has the motor 60?C on the surface. I think it is the weakest point of the roaster. So far I have not better option.

55W cyclone fan :
very powerfull, at 10 it can suck the beans and cause small chaff fire.

The costs in dollars:
black metal plates : 25
stainless : 120
stainless elbow-pipes : 45
motor + 50$ shipping : 56
SSR relay : 17
elastic junction : 12
heating : 85
bearings : 25
cyclone fan : 70
(Raspberry Pi : 50)
PT1000 probes : 50
cyclone fan regulator and cooling fan : scrapped parts
cart wheels : 50
other build material: home depot

total costs: 600-650 $, not bad for me and the whole result .

The result is better than I supposed.
+better uniformity of roasted coffee than from Gene.
+better taste of coffee than from Gene
+better possibility to change profile of roasting
+very easy in operation including cooling than with Gene (manipulation with drum by filling and dumping, cooling)
+clean, by dumping beans is no chaff there (contrary to Gene)

-heavier , bigger (only the roaster 25kg, without cyclone and cooling)
(negative only for those with small rooms)

Now only a few finishing operations (exhaust pipe for winter ? using indoors, ...)

Usually I roast two batches of 400g. Preheating is in 15 minutes completed, cooling down is longer.
danst attached the following images:
3290.jpg 3270.jpg 3220.jpg 3200.jpg 3140.jpg 3130.jpg
What could I do better ?
Better motor and do the same thing neatly.
For larger loads can be heating about 1800-1900W to keep the roaster in range for normal home usage (10A normal home plugs )

The cooling fan is too much powerful, I slow down cooling with separators.
The cyclone serves as cleaning device too.

I give here some information of behavior in espresso roasting. Pre-blend 400g greens.
100% of heating means 1600W.
danst attached the following images:
rst0.jpg 3400.jpg 3440.jpg 3250.jpg 3370.jpg 3160.jpg 3360.jpg
BT and MET on the screen are current values, the roast is stopped, cooling(of the roaster) in progress..
The result, blend with 60% soft dry beans..
danst attached the following image:
The roaster looks great. It should do a good job. Its amazing the level of craftsmanship we are seeing on builds now. Have fun roasting.
All you need in life is ignorance and confidence, and then success is sure. Mark Twain
Nice work on the roaster. It looks great!

It is nice to the the raspberry PI used, I think it has a lot of potential as a built in roaster micro controller. Did you write software for logging?

I got the PI working a while back running RoastLogger as a test.

KKTO Roaster.
I also want to separate the exhaust vent from the loading path, it seems simpler and easier to build. I wonder why more designs don't do that.

Really nice work!
Yes, I wrote SW for this roaster. It can be done with triac or so, but RPi can do the same thing and to monitor temp probes. Since it cannot handle analog PT probes, it must be converted first. To write SW (C#) is easy, although I?m no programmer, but to make a windows program usable for RPi is a lot of work. It make log files, I use this no more, rather I write protocols as above with power and air values. Automatic regulation is with RPi possible too, but I prefer manual control.
Now, the roasts are perfect and repeatable. I use till FC lower airflow, and for 400g beans only 70% of power as maximum. No bad influence on uniformity of beans. I am very happy with this roaster.

Interesting can be, I have the shaft (12mm) fixed in the rear bearing. The front is loose. I can very easy -only 4 nuts - carry down the front face, when needed. The gap between drum and front plate(about 1mm) is adjusted by preheated roaster. No problems with function.
Now, with lower airflow, more chaff is in tray. With extension pipe and cyclone is perfectly removed.
Danst, what is better about the coffee from your roaster vs Gene Cafe? Are you completely satisfied now with your coffee taste, there's nothing left to improve? :)

Also, do you think the chaff cyclone needs to be as large as it is? Could you get by with a smaller diameter cyclone? All the cyclones I've seen are quite large and probably there's a reason.

I'd also like to see a larger picture of the front, so I can see the door design. I'm partly just curious but also looking for ideas?

Nice work!

Mike, my roaster is a compromise between ideas and possibility to do it with limited skills, tools and materials. The door design is very simply and in fact "provisory" , but it works, has no problems and I have no need to make it better.
Now I think , the roaster is not so ugly, more likely "vintage" , from 1m distance.

About chaff cyclone, I believe, smaller is possible, it is limited with fan power and in my case with 12 cm diameter of high temperature fan. The height can be less. But I have no computations, I only tried it so. In my case, I am glad , not for roasting but for cleaning all after, when all can be sucked out. My initial air flows was affected with Gene experience. But less is better. For 10 sec after stopping I go full fan power to suck the rest of the chaff from drum. It is about 40% sitting here and after that is no chaff in coffee. With Gene and soft beans were the chaff rests too big.
In case of solid drum, the cyclone is much more important, it is a need, I think. Filters are complicated with maintenance.
Comparing results with Gene? It is as my previous coffee maker - Gaggia Classic.
It was not bad, but my Bezzera 09, with preheating mod, with two PIDs and absolute temperature stability is way better and easier for use. And the shots are improved and strictly consistent. With Gene, after changing my sort of Brasil beans, I was not able to find satisfying profile. Pre-blend roasting was not satisfying. Now, its no problem. The coffee is better in uniformity, smoother in taste. I do not use the color for determination of degree of the roast. I only hear the cracks and watch BT rise (and make heating and airflow adjustment to follow the profile). I can see , there is no deviation in quality of roasts. I must say, roasting with Gene is as driving a car with only one eye, or even blind. Though, someone can master it. I rather learned from professional drum roasters and their art of profiling. With Gene it is impossible. A good German source: http://www.quijot...stprofile/ , http://www.quijot...t-roesten/
Engilsh-only roasters forgive.
danst attached the following image:
Hey Danst, thanks! Yes your dump door is simple but at least you know how to weld. I'm getting by with no welding ability so far, and it's tricky attaching the dump door to a shaft in just the right way.

The big improvement I made to my coffee lately was to make it sweeter, but I know I can do better because I get hints of amazing taste.

Your description of how you control the roast is exactly how I feel. I don't go by color either and having control over air and heat allowed progress that was otherwise unattainable.

For chaff I may try a small Thien baffle and see how that works. I'm just starting my (2nd roaster) project so it's going to be a while.

Keep us updated!


danst wrote:
What could I do better ?
Better motor and do the same thing neatly.
For larger loads can be heating about 1800-1900W to keep the roaster in range for normal home usage (10A normal home plugs )

I am extremely impressed with this roaster. You've got all the functionality of the best commercial drum roasters and it's got sex appeal!

I can't think of anything you could do better unless you ever feel it could use more heating capacity with the largest batches you'll be running through it. You have features I wish I had included in my 1 lb easydrum roaster and might add down the road.

You've set the bar high for anyone building future drum roasters!

1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
I believe, that a small commercial roaster copy is the simplest way to make it good and easy. Only to have a possibility of good workmanship. It is tricky with stainless , but I am glad I did it. The wrong idea of preheating of incoming air is the reason of only 1600W main heating. For me it is powerful enough, I make two 400g batches weekly. But the drum (7x7 inches) is capable of at least 500g, for it is better to have 1800W or more heating with greater length . Not to go over 5W/cm2 (30W/sq in)-surface of heating- for decent lifetime of it. And for smaller batches can be the power reduced. I have the crucial parts made good for function but not all for good look. This requires much more time and patience I don?t have.
I added an old PC monitor for better possibility to control the roast. The SW is changed, now I have 30 sec table of BT rise to record. The step is 0.5 C. Of course I have an instant preview of 30 sec BT ramp, and now, with bigger screen a graphical progress of set and real BT. Both in 3 sec segments. I still do not care ET, the profile is based on BT progress . The heating power is time programmed, I can change it any time manually. The fan speed changes are only hand-operated . As the fan has different rpm at the same adjustment in different heat conditions , I added a Hall sensor to fan and another thread in SW to monitor, and record correct fan rpm.
The electronic logs are better , I have no stress to record roasting process any more. Only press a button for cracks and the end. I still do not use the tryer , I see the coffee first by cooling. Watching BT is sufficient.
I can say, the building of a drum roaster is completed, I see no reason to improve it more, perhaps to control the fan digitally. After 6 months of experience with this roaster, I am happy and I make now larger batches, about 550-600 g with maximal heating power of about 1300 watts . I can say, this roaster can handle 150 -700g greens without problems. I prefer longer espresso roasts with 5-7 minutes of drying phase and 16-17minutes total. In fact , it is hard to make the coffee bad, in opposite to Gene Cafe.
With batch size changes only preheating temperature and ramp in drying phase. The rest of profiles is the same, of course, the power changes with batch, BT ramp remains.
danst attached the following image:
Very impressive roaster and wonderfully executed! I'm amazed at the ingenuity and fabricating skills of people when I see roasters like this. Add the electronics and writing software... I'm blown away!!


I like longer roasts , now with the rest of my Nicaragua SHG I have tried , what is this roaster able to . With 550g greens , 190C preheating temp and maximal used power of 1440 watts : FC in 10.33 , and Full City+ in 14.15 min. I think, it is not bad. No tipping, scorching , no chaff fire. I am sorry about the quality of my photos (handy).
And I wonder how the short roast tastes.
The color of beans is very even and "warm pastel brown" , so as with longer roasts.
With Gene Cafe was the color more blackish, on every roasting level.
danst attached the following image:
So, as I said, the color of longer and shorter roasts is the same. Although the short roast tastes slightly different, as I had supposed, more crema, more fruity, refreshing, good but not top for my espresso. I stay by longer roasts.

Now I have an info about 1600W heating (5W/cm2) life: it is hard to say, says the manufacturer. It is made of the world best materials and the weak point is the insulation inside. By close to 400 deg C, 100% use, without air circulation(not slow), it should keep 500, maximal 1000 hours of use.
For my 1 hour/weekly with maximal 80% power it can keep 5 or 10 years. Im glad to hear it. I will by another one, for all cases, and forget about it.
Hi danst, I'm extremely interested in your project - see that you are from Czech Republic - me too :) can you please send me contact to you I would like ask you for some details to build roaster like yours :) czech language will be better for me.. many thanks :)
A short video about cleaning chaff with cyclone fan and a pipe adapter :
On my website I have now a conversion of my espresso machine fast completed.
After more than a year of using it :
With the true drum roaster it is possible to manage the roast, much easier than with Gene Caf?, I used before. Very good results.
I prefered longer roasts, I couldn't say why.

E.g. Brazil 12.30-13.00 for FC was OK. With new coffee very bland. The same good results now I have with FC time 10-10.30
I have bought a cheap (~ 20$ ) Digital Grain Moisture Temperature Meter Tester from China. Now I know the reason : A humidity of new coffee is 9% , compared to old with 13%. ("old" is meant bought earlier)
The "drying phase" length must match the moisture of coffee. The rest of the roasting profile can be the same - it must match the bean an the taste target.
Very simple.
danst attached the following image:
The settings of the tester is 1 - wheat , it must be measured with pressure on the beans and probe.
A lot of time is passed...
The wrong ideas can be right after some time...
I made the additional 300W air preheating, as was earlier planned.
I use higher temperatures for preheating of the roaster. The sensitive beans can last out higher temperatures, only the heat must be added at the beginning. I had small defects with brazil beans by 180C preheating and roasting time about 15 minutes, with 210C and roasting time about 11 minutes - no problems, beans without any defects.
For better repeatability of roasts was my SW solution (in console) not ideal.
But the idea of Raspberry as a part of the roaster, without connecting to PC, this was right. I made new SW, in this case in Rpi native language : PYTHON.
The functionality is now as by other usual solutions. I can now compare multiple profiles.
I can use only small numeric keyboard and operate this as "roaster component". I like this simplicity, I am tired of mouse , this is my everyday job routine.
For smaller roasts I can switch off air preheating, and in every roasts it is switched off automatically at 170C - for possibility of cooling after FC.
The key for roasting is to achieve about the same preheating values - BT,MT,MET and then is repeating very easy.
For indoors conditions I have improved air exhaust. It was necessary, I make now more coffee, about 4kg/month.
danst attached the following images:
679.jpg ugly2.jpg ugly.jpg 671.jpg
It's good to hear you've got excellent control over your roaster. It looks like the preheating stage has helped get a jump on initial high heat needed at the start of a roast.

Have you found a favorite profile with this roaster that works well with most coffees intended for non-espresso brewing?

1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
I have the same feeling, the preheating helps. I make the coffee now closer to Rao?s principles. It can be right too, a perforated drum needs shorter roasting times. The profiles above are good for coffee with minimal acidity. I like it so for espresso. I usually mix (not overly traditional blend) 75% of this brazil ,up to 222-3 C, dumped just on hint of SC or close before , with 25% of colombian or nicaraguan coffee . This is roasted to 226, 20 sec in SC. It still preserves acidity and has no oils on the surface.
I drink moka or drip only on vacancy time, where I can?t make espresso.
For this purpose, I make nearly the same coffee, I dump it about 5 degrees before. Sometime I give this coffee to a friends and they are very satisfied with it. But this is very subjective classification.
I know, when SC is about 223 C, the recommendation for non-espresso is dump about 208 - 210 C, but this works not for me.

I believe too, there is no universal profile (I?m searching for), it depends on bean, roaster and taste preferences.
My profiles now generally meet 150 at 4 to 4.30 , FC 8-9 , end on 11-12.

Till I detect, it is wrong.
Jump to Forum:

Similar Threads

Thread Forum Replies Last Post
DRUMTC Roaster Turbo Oven Roasters 12 05/13/2023 9:17 AM
3kg Drum Roaster Build Using 40qt Stock Pot Building a Coffee Roaster 3 04/19/2023 10:46 AM
Is vacuum cleaner motor ok for fb roaster? Fluidbed Roaster 35 04/17/2023 11:22 PM
Aliexpress Fluid Bed roaster 300g Fluidbed Roaster (Store-Bought and Commercial) 35 04/11/2023 8:27 AM
All About the Bubble Bed Roaster Fluidbed Roaster 64 04/08/2023 10:11 AM
Homeroasters Association Logo, and all Content, Images, and Icons © 2005-2016 Homeroasters Association - Logos are the property of their respective owners.
Powered by PHP-Fusion Copyright © 2023 PHP-Fusion Inc
Released as free software without warranties under GNU Affero GPL v3
Designed with by NetriX
Hosted by skpacman