There's too many variables to explore successfully. So my current practice is very simplified - I run a nearly constant MET between 250C and 260C (480F to 500F), which is as tight as I can get the control. If I want to roast slowly, I run at low heat and low airflow, if I want to roast fast, I run it at high heat and high airflow, but always at this MET

So now I dry very low heat and air, ramp to the first at max heat and air, and finish at medium heat and air, twiddling the fan to keep the temperature around 260C.

I do not know if this technique carries over to other Probat style roasters, or if this is peculiar to the Quest. If this technique holds up for me, I'll PID the heater to the MET, and will control the roast speed by twiddling the air knob.

Just as another data point, my current methodology is a "3 and a half ramps". Bean loads are 350g.

In terms of duration (all temps are BT) :
Ramp 0.5 - drop-in to 250F : anywhere from 1:30 to 3:30. Heat is on full blast.

Ramp 1, aka "Drying" (250F-300F): 3:00-4:30 - typical is 15F/min

Ramp 2, aka "Into 1C" (300F-1C, ~400F): 3:00-3:30 - typical is ~33F/min rise, but I taper at ~385F to slow the intro to 1C.

Ramp 3, aka "1C to end" (400F-end): 3:00-5:00 - typical is 7-8F/min, which puts espresso roasts that end right around 2C at 430F closer to 4:00 in duration.

I'm still playing a little with a slower 0.5 ramp, but am starting to doubt the benefits. After all, why would I _want_ to take longer?

chang00 wrote:Are the bean mass probe and MET reliable readings that can be depended on for reproducible roasting?

I think this may be device dependent. For my roasting method, I would be in the camp of "yes, BT probes can definitely be dependable for reproducing results".

-s.

I always roast at least 2 batches when a new bean arrives. The first being my best guess and the 2nd depending on how the bean acted and often either pushing or easing or just straightening out where I was off on the first trip.. Tweaking the variac quite a bit keeps me with the roast.