Just thought Id get my latest failure off my chest, and maybe get some others thoughts on temperature measurement and paddle issues.
Ive been roasting for a while with a simple heat gun and bowl setup, as shown in the first photo. It worked very well, except for unstable temperature readings - temps jumping up and down according to how I stirred the beans, and how often the probe was briefly exposed to a direct heat gun blast, or an ambient wind gust. FC and SC temps varied too - one day FC would occur at 185, the next day at 165, or so the reading said. In the end I didnt put much trust in the temps, and got good results based on gun height and the feel of it. But I thought ... there must be a better way.

Unable to find a 2nd hand breadmaker that didnt need modding, I had an idea ... for coffee roasting purposes, isnt a breadmaker just a tall tin with a motorised paddle? Couldnt I achieve the same function with a stainless steel biscuit tin, a motor from Jaycar, and a home-made paddle? The following photos shows the result, where it differs from my previous setup (with closer view of paddle and probe, minus tin).


The tin has two holes in its bottom - the temp probe sticks up through one at the edge, and the motor axle sticks up through the centre. The scrap metal paddle sits snugly on the axle, but easily comes off when the tin is lifted upwards at the end of the roast. The geared motor axle spins at 12 rpm with enough torque to stir lots of beans ... perfect, I thought.
Maybe the higher tin would keep outside wind gusts away from the beans, and together with the deeper and more evenly stirred bean mass it would lead to more reliable temp readings. And I wouldnt have to stir with a wooden spoon! I was feeling pleased with myself and my "simulated corretto" ... until I used it.
Ive just produced my most uneven roast ever. The paddle seemed to stir well, yet some beans still looked green while others were at CS7. Some were still starting FC when others were well into rolling SC, a few on the brink of burning. Even after removing all the charcoal beans, the rest ranged from pale CS5 to dark CS11. Sure, the temp readings were more stable, but no more accurate than before (FC at 160?).
So not all progress is forwards. Im now thinking what I could change, or whether to go back to the bowl and wooden spoon, and Im wondering:
* Has anyone else had problems with temperature readings varying due to factors other than the actual bean temperatures?
* Do you need a certain minimum bean depth to get an accurate temperature? (I roast no more than 230g at a time - my weekly consumption - but I know others roast much more in corretos)
* Does the paddle need to be a particular shape/size/angle to get thorough mixing? (The breadmaker paddles Ive seen were basically flat rectangular surfaces, like my home-made paddle)
On the bright side, if someone wants a strikingly mottled roast with raw notes and carbon aftertaste, I can produce one!
Ive been roasting for a while with a simple heat gun and bowl setup, as shown in the first photo. It worked very well, except for unstable temperature readings - temps jumping up and down according to how I stirred the beans, and how often the probe was briefly exposed to a direct heat gun blast, or an ambient wind gust. FC and SC temps varied too - one day FC would occur at 185, the next day at 165, or so the reading said. In the end I didnt put much trust in the temps, and got good results based on gun height and the feel of it. But I thought ... there must be a better way.

Unable to find a 2nd hand breadmaker that didnt need modding, I had an idea ... for coffee roasting purposes, isnt a breadmaker just a tall tin with a motorised paddle? Couldnt I achieve the same function with a stainless steel biscuit tin, a motor from Jaycar, and a home-made paddle? The following photos shows the result, where it differs from my previous setup (with closer view of paddle and probe, minus tin).


The tin has two holes in its bottom - the temp probe sticks up through one at the edge, and the motor axle sticks up through the centre. The scrap metal paddle sits snugly on the axle, but easily comes off when the tin is lifted upwards at the end of the roast. The geared motor axle spins at 12 rpm with enough torque to stir lots of beans ... perfect, I thought.
Maybe the higher tin would keep outside wind gusts away from the beans, and together with the deeper and more evenly stirred bean mass it would lead to more reliable temp readings. And I wouldnt have to stir with a wooden spoon! I was feeling pleased with myself and my "simulated corretto" ... until I used it.
Ive just produced my most uneven roast ever. The paddle seemed to stir well, yet some beans still looked green while others were at CS7. Some were still starting FC when others were well into rolling SC, a few on the brink of burning. Even after removing all the charcoal beans, the rest ranged from pale CS5 to dark CS11. Sure, the temp readings were more stable, but no more accurate than before (FC at 160?).
So not all progress is forwards. Im now thinking what I could change, or whether to go back to the bowl and wooden spoon, and Im wondering:
* Has anyone else had problems with temperature readings varying due to factors other than the actual bean temperatures?
* Do you need a certain minimum bean depth to get an accurate temperature? (I roast no more than 230g at a time - my weekly consumption - but I know others roast much more in corretos)
* Does the paddle need to be a particular shape/size/angle to get thorough mixing? (The breadmaker paddles Ive seen were basically flat rectangular surfaces, like my home-made paddle)
On the bright side, if someone wants a strikingly mottled roast with raw notes and carbon aftertaste, I can produce one!

Comment