FZ94 warmup, profile design and roasting Guatemala Las Delicias

It's not easy to estimate the amount of energy inside the warmed up Coffee-Tech FZ94 machine so sometimes after charging the load of beans, the Turning Point is later or sooner than expected, at a lower or higher temperature than planned and then the roast will take longer or go faster than anticipated.

This is, in my view, not necessarily a problem as long as the part after Dry End is largely according to the desired profile with the development time sufficient to get the desired roast color as measured, for instance, with the Tonino device.

To get a warmup situation that is as consistent as possible, I developed a setting of the internal PID of Artisan which regulates the airflow to precisely keep the BT probe at a certain temperature. I aim for the other probes, Drum Temp and Exhaust Temp, to be close and stable near that BT value:

Warming up the Coffee-tech FZ-94 roaster

You see in the screen snapshot that the target BT (brown line) is at 155ºC. I start with all 3 elements on and at 10 minutes, as DT (black line) goes through the 155ºC mark, I switch off one element. As the RoR (dark blue line) levels off and BT reaches 155ºC around 14:30, the airflow (light blue line) is increased and for 10 minutes the machine is kept very stable by the Artisan PID feature. ET (blue line) is practically the same as BT here.

These are the PID settings that I use:

And this is the setup of the slider. The lowest value of the slider still has 55% of the total fan speed so no hot "bubble" can build up beneath the drum. If such a bubble is allowed to build up, it would belch upward when airflow is increased and then for a while, the opposite effect (warmup) would be accomplished and not a cooling (or actually less heating) effect of airflow increase.

Slider values:

Below is the design for the next roast to start right after warmup is completed:

And this is a first result, roasting Guatemala Las Delicias by Manolo Muralles:

Some minor shifting left-right has been applied during the roast to align the BT curve with the plan. One element shut off shortly before FC. Could maybe have been shut off later, avoiding the slight slump in RoR which is exaggerated in the graph by the blue RoR line. Airflow around FC was not very high so there was margin enough to correct some surplus heat from the third element. Second element shut down a littel past 14 minutes and the last one at about 15:30. Roast color after completely cooling down 30g of beans, chucking 10g through the R120 Compak to get any other grinds out and grinding 20g at extra fine setting, was Tonino # 91 which is within my target of 95 plus/minus 5. I let 17.5g of puck rest for 5 minutes before pulling a first espresso which was not as sharp as anticipated for such extreme freshness. 28g in 25s, a spicy creamy cup, not dry in aftertaste, so a good mouthfeel.

A picture of an espresso made with an earlier roast, Batch 5:

Looks good, is pleasantly spicy but a little dry in aftertaste. Works better with milk and updosed / short as that improves / masks the dryness, but it's clear to me that Batch 5 was roasted too fast (a little over 15 minutes compared to 16:43 of Batch 10).

PS next day, roasted along the same method for Batch 11:

TP was at 99ºC and 2:03min, practically identical to previous roast. At DE temperature BT was at the planned spot and the 'dotted line' indicating the current direction pointing towards the planned FC point so I activated the PID to keep tracking along that line. FC started as planned, earlier than last batch and I switched off the first element later, when I observed the airflow almost maximised by the PID, at the start of FC. Second element off around 12:45, which I very briefly switched on again just before 14:00. RoR was a bit lower than anticipated and I waited for BT to hit 175ºC before stopping the roast at 16:16.

Roast color measured with the Tonino practically the same as previous batch, so nothing unexpected.