Homeroasters - Home Roasting Coffee Community » BUILDING A ROASTER » Drum Roasters
Who is here? 1 guest(s)
Page 1 of 2: 12
 Print Thread
Another electric 1kg roaster
danst
Top  #1 Print Post
Posted on 05/06/2019 11:44 AM
This is my old Ugly roaster, in fact up to 700g greens :
https://forum.hom...rowstart=0

After almost 5 years of roasting I decided to make a new one. 50% larger.

Why ? I?d like to have a roaster that is capable of roasting in batches with 500-1000g result.

I have positive experience with Ugly 600, so I use this model and I change it a bit.

The changes :

Drum 219mm x 190mm (dia x length) , 3 mm stainless. (Up to 1200g greens ?), intended mostly for 630 and 950g batches.
Perforated material is very difficult to obtain, so I use now solid drum design. But with large diameter and perforated back (RV 3-5).
The frontplate not welded- only screws, it can be replaced with another one.
There is stainless only on drum and frontplate (4mm thick at first).
I believe the roaster don't need 100kg...

I need to make the roaster group shorter (room space reason), so I use a new smaller chaff cyclone.
The fan must be new, with smaller diameter.
I use a new short drum motor with angled gearbox.

The heating is the same , with main part doubled. The total output close to 4 kW.

I use old cooling fan and electronic of Ugly I (3 probes PT1000 and RpI with my SW). I make (I hope) motor powered cooling tray.
With other details -I have no precise idea, some things must be solved on the fly...

I must complete parts and then recycle the old roaster, without big dwell time.

With limiting tools and skills it?d be ugly again, I hope without functional compromises.....
(toy welder and 1mm sheet metal - it?s a pure fun)

Here some pictures of progress:
Edited by danst on 05/07/2019 12:59 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #2 Print Post
Posted on 05/07/2019 2:48 AM
Pictures:
I can?t upload the .jpg files, why ?
Edited by danst on 05/07/2019 5:10 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
renatoa
Top  #3 Print Post
Posted on 05/07/2019 5:06 AM
Edit post, look bottom button "Choose File", browse, select, Save Changes.
 
danst
Top  #4 Print Post
Posted on 05/07/2019 5:13 AM
I make this from yesterday many times , without succes, new reply too.
File type OK, size, number , the same way.
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #5 Print Post
Posted on 05/07/2019 5:22 AM
The last try to upload pictures with another browser...
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
JackH
Top  #6 Print Post
Posted on 05/07/2019 6:23 AM
Make sure your filename is normal and without spaces or special characters. Do not use the "preview post" or "Preview reply" button. It somehow causes problems with the images. You can have a file as large as 10MB.
JackH attached the following images:
variac_web_1.jpg roast_web_1.jpg mod_oven_back_6.jpg mod_oven_web_6.jpg
---Jack

KKTO Roaster.
 
danst
Top  #7 Print Post
Posted on 05/07/2019 8:54 AM
No special characters and spaces, only numbers. I try a to z only.
Nothing.

And now longer name.
And another, It goes now, I must get only one picture and repeat it by editing, although it ?d be max 8 together.
(It was only 7 but failed)
danst attached the following images:
komplet_7.jpg ad_1.jpg ac_1.jpg af_1.jpg ag_1.jpg at_1.jpg
Edited by danst on 05/07/2019 9:16 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #8 Print Post
Posted on 05/07/2019 9:10 AM
And the answer is : the 7. image caused problem, after opening and saving (ungchanged) it goes now.

Sorry, bad luck, it is too much needless replays.
danst attached the following image:
az.jpg
Edited by danst on 05/07/2019 9:19 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
JackH
Top  #9 Print Post
Posted on 05/07/2019 10:55 AM
Odd problem and I am glad it works for you now. The system is Linux based so I usually suggest avoiding Linux reserved characters like / > < | : &.
---Jack

KKTO Roaster.
 
danst
Top  #10 Print Post
Posted on 05/21/2019 2:45 AM
some progress....
danst attached the following images:
p2_2.jpg p3_1.jpg p5_1.jpg p6_1.jpg p7.jpg
Edited by danst on 05/21/2019 10:35 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
Liam
Top  #11 Print Post
Posted on 06/27/2019 4:11 AM
I hope you got exactly what you wanted smile Based on the technical specifications, this should be something big!
 
BobbyS
Top  #12 Print Post
Posted on 06/27/2019 11:40 AM
This is coming along nicely. I bet it is going to work great! Nice job!
 
danst
Top  #13 Print Post
Posted on 07/10/2019 4:29 AM
The time is short, I tested the uncomplete roaster for heat behaviour and drum function.
No heat problems, the drum is OK, but 2 issues appear here :
1. the chute funnel valve is not optimal - some beans drop to air tube
2. the front plate door is small - to empty all beans takes 18-20 sec

The vlave is redesigned already, the door can be with new plate larger. I don?t know if it can be a problem.
I guess, a minor problem. In most roasters it takes about 10 sec, but longer times too. It is a constant and so unimportant. And melange type roast is always partially present.
The thickness of the front plate (4mm) is now the last unknown. In case of a new , 5 or 6mm thick, the door will be larger, when is 4mm OK, the door remains old.
danst attached the following images:
b4.jpg b3.jpg b2.jpg b1.jpg a13.jpg a12.jpg a11.jpg
Edited by danst on 07/10/2019 4:41 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #14 Print Post
Posted on 07/10/2019 4:55 AM
Now, I have finished the first version of cooling tray, with provisory motor: 60KTYZ 50rpm. The right 68KTYZ with 20 rpm is on the way. The accuracy is not my thing, I use a hanging system to accept all odds. The function is perfect with 50 rpm too.
The roaster is now installed on the place.
danst attached the following images:
c3_2.jpg c2.jpg c1.jpg b3_1.jpg b2_1.jpg b1_1.jpg
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #15 Print Post
Posted on 07/10/2019 5:17 AM
Again with incomplete roaster, I tried the probes and roasted some beans, 650 and 950g .
The BT probe was now 12mm M4 PT1000, but the behaviour was not good. It was far away of what I know.
I use next my old 40x3mm probe again. The mass of the roaster is much bigger then by Ugly I and I must adapt to it.
From first impresions, I beleave maximum 1250g roasts are possible.
After that I completed air venting system and monitoring of air pressure with new sensor : +-500 kpa SDP816-500.
All functions are ready, but cooling air and roasting exhaust in one tube - this is false. The cooling air goes by return to branch pipes, I have this not assumed. I try it once again by roasting.
Now, the all electronics is provisory connected and for longer time it must so remain.
I must split up venting and adapt to the new roaster. Then perhaps think out front plate thickness.
But the most is done, I can roast again.
danst attached the following images:
d2.jpg d1.jpg d3.jpg
Edited by danst on 07/10/2019 2:35 PM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
allenb
Top  #16 Print Post
Posted on 07/10/2019 3:54 PM
As usual, great roaster build! Sounds like once you've gotten your exhaust venting worked out you will have a very capable 1.25 kg shop roaster. Keep us updated on progress and also let us know how well you're able to control it.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
danst
Top  #17 Print Post
Posted on 07/11/2019 2:25 AM
Next weeks show all possible problems...

Here is a short video of cooling tray in action, this system with loose silicone (or teflon) lath can overpower all inaccuracies :

https://youtu.be/...
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #18 Print Post
Posted on 07/17/2019 9:06 AM
The problem with cooling successfully solved.
In branch pipes are helping counters made. There is no problem with reverse air already.
But in a roasting session, it must be at least very small flow in smoke fan to prevent smoke to go reverse from cyclone fan - I have no problem with it.
The pictures can show it.

A profile (roasting with only 650g beans) shows, that there is no problem to cool down the roaster (this was in fact too quickly).
I have calculated the mass of the front plate. This large diameter 4mm plate matchess 6mm by North and others, with diameter closer to drum dimension. That means, there is a room to go further, I think 6mm, as the others go with more then 6mm ?
In winter it can be preferable to have more heat storage and the roaster can manage it in summer too.

In the profile the horizontal lines from above :

blue : cyclone FAN pressure sensor
purple : cyclone FAN rpm
red/brown : MET (3x12mm sensor on frontplate outside the drum)
green: BT (3x40mm sensor in beans)
violet/brown : ET (3x30mm sensor in feed chute)
declining: RoR
red: heating

The roaster is now rough ready, I must master new behaviour and then try the thicker plate. But this can wait.
So as a minor finishing.

I can say, the new roaster is better then Ugly I , now it is a true drum roaster. The beans are completely without scorching or tipping, even with shorter roasting time. The difference in between FC and the end is about 5C smaller.
The coffee is more "rounded", I prefer it. ( solid vs perforated ? big mass vs mass? )

I like the idea with only 1 venting pipe compared to 3 ones before.
All this automated cooling is pure fun too.
The "preheating" of the air (acces of air thru warm section conducting or direct from cool ambient) is a good upgrade and the end of roasting can be more easily done.

It was worth the whole work.

Additional info : roaster alone: about 33 kg
cooling tray : 10 kg
cyclone : 6 kg (only 0.7mm plates)
pipes : 3kg ?
carriage : ?

The whole control by Rpi and manual switches, variable drum speed 50-80 rpm.
Stand-alone equipment, no laptop needed.

By the way..... how thick is North 1kg front plate ?
danst attached the following images:
e1.jpg s3.jpg s2.jpg s1.jpg s0.jpg
Edited by danst on 07/21/2019 3:09 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #19 Print Post
Posted on 08/31/2019 3:59 AM
Nobody knows North 1kg front plate thickness??

I can now roast 650 and 950g (1kg) batches reliably, although it needs a small improvement of managing the temeperature, as the example photo (profile) shows.
With the old BT probe and setting of reasonable airflow the temperatures match approximately the values I was used with Ugly I. (FC so as SC, 196-200 / 221-223)
The difference is absolute no scorching, no tipping, even by sensible beans. Probably lower charge temperature.
And second, the DTR for espresso does not require more than 25% - better developement.
One thing is worse - FC is not so audible.

I decided to try a new front plate. 7mm thick, a bit smaller in dia, with greater door and without glass window - it seems useless, I have no advantage with the sight glass. (only the first validation of "yellow" temperature and probe reading)
Why 7mm ? I think , I have only 2.5mm drum and for winter time I?ll need more stored energy.
I can find, it is wrong, then I take the older one again.

I have a bit upgraded RoR drawing, the values are now much more actual, the time lag is now about 6 seconds.
I must still replace the 50rpm motor in coolling tray. The speed is too high for comfortably testing of the temperature with the hand.

And even one thing is not a great success. The new smaller chaff cyclone is not so efficient as the old one. More chaff goes thru, but it is not a problem for me.
danst attached the following image:
p1.jpg
Edited by danst on 09/02/2019 6:22 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
allenb
Top  #20 Print Post
Posted on 09/07/2019 5:55 AM
One would think that simply increasing the size of a roaster while keeping most design elements the same would allow the roaster to perform with close to the same characteristics as the smaller original but is never the case. It always requires alterations until the roaster behaves correctly.

The log you posted is interesting to me as it's almost identical except for development time to a very recent roast of mine from last week with a Costa from Happy Mug which turned out to be exceptional. This coffee preferred a quicker, 1.5 minute development in order to pull out it's complex flavor notes. Going beyond 2 minutes caused most of the notes to be very muted although still tasted good.

Another side note is this coffee had a particular underdeveloped green taste if started with a preheat drum temp of less than 375F and if starting the roast with too little power. Best results were with RoR high enough at start of roast to hit 200F by 1 min 45 sec.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
danst
Top  #21 Print Post
Posted on 09/15/2019 12:36 PM
Yes, there is a difference , but I know, I have made improvements, not only larger roaster.
The passive preheating tunnel is good for better energy effectivity, all parts with problematic maintance are overworked, it will be working better, some parts are better designed (I have avoided some earlier mistakes).
I have slowly practice with them, the roaster can without problems make 1kg batches very fast - FC can be in 8min.

But this is now history again. I must learn new configuration of this roaster. The new 6mm (7mm was not at disposal) front plate and greater door is from now in action. The photo shows the old bent proofed BT probe and 2 short for ET and MET.
I think for in-drum temperature measurement was 12mm probe too much influenced with front plate, I changed it with 30mm one. For MET, I think, the short probe can better project thermal status of the whole roaster. And this is needful condition for batch roasting, especially with electric heating.
And I can be wrong, of course.
danst attached the following images:
a2.jpg aa1_4.jpg
Edited by danst on 12/31/2024 4:11 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
Erichimedes
Top  #22 Print Post
Posted on 09/20/2019 10:38 AM
Danst, your roaster looks fantastic!

I am about to build a 1kg drum roaster, and would like to use electric heating if possible.
You mentioned having about 4000w of electric heating. I'm curious how much of that wattage you use on a 1kg roast? I know you also preheat your incoming air, which I think is very smart. I'm wondering if I can get by with 5kw of heating and no preheating of the incoming air.
 
danst
Top  #23 Print Post
Posted on 09/23/2019 4:15 AM
Some roasters use 2.5kW /1kg beans. They have thick front plate and drums mostly perforated. Others have 4.5-6 kW/1-1.5 kg and the same built fashion as gas roasters with solid drum.
My has 3.8kW (as I used spare heatings from earlier) . Now with 6mm f.plate is the behaviour as I presupposed.
(Still I have an idea to try adding more mass, simply what happens by it)
The use of power corresponds to preheating level. 80% - 10% during the roast mostly.
I think, we are at one, the roaster is not only the heating. All parts relates together.
5kW / 1kg is fine.
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #24 Print Post
Posted on 10/02/2019 11:23 AM
The new slow 20 RPM motor with high torque is in action. Now it?s safe to try the bean temp with the hand.

It was a pity to throw out the old 4mm front plate. I cut it and added about 15% mass for heat storage.

Now is the roaster completed and ugly again.

The tehnical behaviour is very good and the result is up to operator. And that is how it should be.
danst attached the following images:
aa1_3.jpg aa2_3.jpg ab1.jpg b16.jpg b1_2.jpg b2_2.jpg b4_1.jpg
Edited by danst on 10/03/2019 8:58 AM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
danst
Top  #25 Print Post
Posted on 03/01/2020 12:56 PM
Now, despite of cold winter ambient, I'm happy with the roaster. I have better control/experience on it, the airflow is better adjusted, not so much as in earlier attempts. I still make minor mistakes with heating, but the results are good.

I also added 4. temperature probe on the front plate, this is useful for checking of preheating. (MT, ET, BT, FT on the picture)

Here an example of 900g medium/dark - full city roast (some people still like this, including me BBQ grill).
danst attached the following image:
obr.jpg
Edited by danst on 03/01/2020 1:04 PM
1000g Ugly roaster, Bezzera BZ09 PID, Fiorenzato Doge 63 stepless, Mazzer Jolly(Mestre), T. Moccamaster.
 
Page 1 of 2: 12
Jump to Forum: