For some time now I've had plans to build a coffee roaster out of a small electric oven with a rotisserie and last week I decided it was time to get to work. The first problem I encountered was making a drum. It would be difficult for me to do that with the limited amount of metal handling experience I have. Since the Behmor was my big example I decided to contact the local Behmor importer and to my luck they had a used drum they wanted to sell for 15 euros. With the help of some Home Baristi I found out that the Behmor drum would fit the oven I had in mind and decided to order the drum and 2 kgs of green beans as well as a motivation.

The Behmor drum weighs 400 grams, is made out of wire mesh and has clips to hold shut one of the side walls. It has nice blades inside that will push the beans outwards and back inwards as it spins. The dimensions are 12.5 cm in diameter x 26 cm in length with 1.5 + 1.7 cm axles.





One side has a square axle. The other side has a round axle.



After the drum arrived I did some research on electric ovens with a rotisserie that would fit my drum. It lead to 4 brands:

• Bestron DGL26, DGL30
• Severin TO 2017, TO 2024, TO 2033, TO 2036
• Steba KB 23, KB 27, KB 28
• Tefal OV1000/Tefal Compact Grill

I found a fairly priced Tefal Compact Grill and 120 km and 25 euros later I was the happy owner of this nice little oven. At home I immediately unscrewed the cover and was happy to find out that the technique used was very simple. It has 3 independant heating elements: 2 above and 1 below.




The right hand suspension of the rotisserie:


My challenges would be:

• The rotisserie was placed diagonally so the drum would scrape into the side wall.
• The drum was shorter than the inner dimension of the oven so the axles would probably need to be lengthened.
• The square hole to place the rotisserie in looked smaller than the drum's axle.
• The motor only ran in the 'rotisserie programm'. In that programm only 2 of the 3 heating elements would light up and I would like to use all 3 at once.
• I wasn't sure if the motor would spin fast enough.
• The oven had a thermostat but I didn't trust it and would like more control over the temperature.

Step 1: Moving the motor to the side wall
The motor was mounted on a separate bracket and looked easy to move. I unscrewed it, drilled one large hole in the side wall for the axle and two small holes to hold the motor and in 15 minutes the motor was moved to its new destination. Luckily the square hole in the axle turned out to be the right size for the drum and was just a little bit rusty.





Step 2: Making a bracket for the suspension of the left side of the drum
In the left side wall I also drilled 2 holes and used a metal strip with a V shaped cutout for the drum's left round axle.



That's all that was needed for mounting the drum. I'm very glad I didn't need to change anything to the drum's axles.



Step 3: Having the motor run in all programms
Using a multimeter it was easy to determine the engine was only supplied power in the rotisserie programm. Using a piggy back connector I clipped the motor's cable to the power switch so it would run in any programm, independant of the thermostat. It's the white wire leading from the motor to the left.


Step 4: Test run
The oven passed its first empty test run. No fuses were blown and no other calamities occurred. Furthermore the rotation speed was high enough. The drum made a full turn in 30 seconds. The heating elements above and below lit up nicely.



Step 5: First roast
Using 3 small 50 gram batches I did a couple of cautious test roasts.


The coffee was an India Monsooned Malabar.


First batch using 3 heating elements. First crack at 7 minutes. Second crack almost immediately. The result: burned, oily, unevenly roasted. Nasty!


Second batch using just the 2 upper heating elements. First crack at 8.5 minutes. Second crack 1 minute later. The result: also burned, oily, etc, etc. Started to get annoyed


Third batch using the upper 2 heating elements. First crack in 10.5 minutes (the oven had probably cooled down more after the second attempt). I quickly opened the door a bit. Second crack: uhm... didn't happen. I eventually stopped the roasting after 25 minutes. The result: too light. But at least the beans had a nice even color.


BTW: Masses amounts of smoke came out of the oven when opening the door after first crack! Strange that I coundn't see or smell that when the door was closed. Luckily the oven was placed under the kitchen's ventilation hood.


This is a picture of the very modest amount of chaff after roasting these clean beans.


Here is a video of the first run. Please pay attention to the sound of first crack .


Step 6: Building some kind of temperature regulation
That's the part I haven't done yet. Having PID'ed my Gaggia Classic recently (text in Dutch) I realize it won't be hard or expensive to build a PID into this oven. The only challenge here is to find a PID that supports 2 set temperatures so I can easily lower the temperature at first crack.

Food for thought

• I was lucky in finding a usable drum.
• I expected more fuss mounting the drum and I guess I got lucky here too.
• Three heating elements is way too much in a closed oven.
• I would really like a small lamp inside the oven to be able to keep an eye on the roasting process. My led flashlight sort of works but it isn't great.
• The motor is attached to the side wall and gets hot. I touched the original motor bracket while roasting and that one also gets hot. I'm hoping the motor will last.
• It would be nice if I could do something to get rid of the smoke. The oven is supposed to be placed on top of my microwave but the ventilation hood doesn't reach that far.
• During the third roast I noticed the drum doesn't always turn in the same direction. I recognize this behaviour from the orange press but here I cannot appreciate it because the blades in the drum need rotation in one specific direction. If anyone has any ideas on how to lock the rotation direction I would be very happy to hear it.

I hope this report inspires others and am open to any suggestions .

PS: I submitted this report to the Dutch barista forum a couple of days ago.

Arne

The lower elements seems dangerous for the potential of a chaff fire. With your beans right above would add fuel to a chaff fire.

atjong wrote:[ I realize it won't be hard or expensive to build a PID into this oven. The only challenge here is to find a PID that supports 2 set temperatures so I can easily lower the temperature at first crack.

Cool project

I'm wondering why you would want two set temps, and want to lower the temp after first.
You don't ever want to lower ET, it should always go up. It's just that you want it to slow down the ramp rate after first, compared to getting to first.
The PID needs to support at least a few different ramp rates, so you can have different ramps for different stages of the roast.

If you could figure out how to get a thermocouple into the bean mass, so you can monitor Bean Temp, and then put a thermocouple by the heaters, to monitor environmental temp, then you would be set. I use ET for PID control, and then monitor BT to help make decisions on when to stop the roast and make sure the roast goes ok.

Randy

2 rpm is way too slow for the drive motor. That explains the uneven roasts in the pictures.

First, congratulations on your ingenuity.

As previously pointed out, the rotational speed is very slow; unless you can speed it up, you will have trouble getting results that are any better than oven roasting in a pan. I would certainly not let the chaff accumulate from roast to roast, because as Ed pointed out you would be inviting a fire.

Smoke production is going to be proportional to the amount of coffee that your roast and to the degree of roast. Because this is not a purpose built roaster with an exhaust pulling the smoke in one direction, no matter what you do with this homebuilt roaster you will need to be prepared to deal with the smoke. Unless you can get the thing right under a smoke hood with a strong fan, using this will prove to be a very smoky experience, especially if you roast more than 150g or so in each roast.

ken

farmroast wrote:The lower elements seems dangerous for the potential of a chaff fire.

I'm not using the lower one anymore, thanks. Just the upper two.

randytsuch wrote:I'm wondering why you would want two set temps, and want to lower the temp after first.
You don't ever want to lower ET, it should always go up. It's just that you want it to slow down the ramp rate after first, compared to getting to first.

That's what I believed to have read to be common. To slow down the roasting process after first crack. But I'm very much of a rookie to this and reading your comments I realize that I was wrong. Thanks for the advise. Next step is to find a universal PID that support different ramp rates. I think that wil be even more of a problem .

randytsuch wrote:If you could figure out how to get a thermocouple into the bean mass, so you can monitor Bean Temp, and then put a thermocouple by the heaters, to monitor environmental temp, then you would be set.

I don't think I want to go that far at this moment. That will cost some major changes to the drum. I would like to stick to enviromental temperature for now.

Gime2much wrote:2 rpm is way too slow for the drive motor. That explains the uneven roasts in the pictures.

Probably, but still the 3rd and 4th roast were pretty even. The big difference between the uneven and the even roasts were the rotation direction. In the first 2 roasts the direction was wrong. Here's a picture of the beans from the 4th roast:

another_jim wrote:First off, these roasts are easy to judge: a first pops of the second roast should have white splits, and a smooth, fully expanded, even colored, uncharred surface. If it can do this, it has covered the basics, that is roasting the beans without dehydrating, charring or smoking them.

Thanks for the advice. Currently I only have Cuban and a Peruvian green beans available but I'll try to find what you just described in my roasted beans.

Ken Fox wrote:Unless you can get the thing right under a smoke hood with a strong fan, using this will prove to be a very smoky experience, especially if you roast more than 150g or so in each roast.

While still only in test run phase I still was surprised to see my 4th 50 gram batch barely created any smoke. With 29.5 minutes it was a long roast because I only used 1 of the 3 heating elements but still I expected more smoke.

Very nice to read your experieced views on this Jim. Thanks, I really appreciate it.
Tonight I tried a small 50 gram batch of the Peru washed Arabica.


I roasted them just a few seconds into the 2nd crack. This took 18:40 minutes using the upper two heating elements. This is a little longer then I'd like it to be but this is probably because I fiddled around with the oven's thermostat. Once I've PID'ed this thing I guess I'll have more 'regular' roasting times. There was only a small amount of smoke and apart from a few defects the roast looks pretty even to my rookie eyes. The results:



I'm very curious of your opinion on this.