Air Roaster vs Gas Drum Roaster Insight - Page 2

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
Rush (original poster)
Posts: 90
Joined: 9 years ago

#11: Post by Rush (original poster) »

false1001 wrote:Looking at these profiles I would suggest:
1) Charging at a higher temp
2) Trying to reduce the time difference between the drying and Maillard phase. They don't have to be the exact same time, and some coffees like longer drying phases, some like longer maillard phases, but a 2 minute difference is a little extreme.
3) Aiming to drop at a slightly higher temperature

Of course, all roasters are different. Charge size and bean density have large effects on FC behavior as well.
Thanks for the input. I tried to find a profile that fit your description, I have tried all sorts of ideas in the past. This one is pretty close, it is 22.6% of rated maximum charge mass. The result is pretty good, probably least defective tasting. Still a hint of burnt/toast, but not grass like a roast with lower finish temp.

User avatar
Almico
Posts: 3612
Joined: 10 years ago

#12: Post by Almico »

Rush, I understand the desire to "upgrade" equipment, but I'm not convinced it's the Cafemino here.

For one, I'm not in the "gas is more responsive and therefore better for roasting coffee" camp; especially in a drum roaster. Power changes in roasting coffee should be few and subtle. If you have to make a drastic power change during a roast, it's to recover from a mistake or lack of a roasting plan.

Nimble air roasters, like the Artisan 3, can be a crutch that does more harm than good. When I was using mine, I had very little roasting plan at all. I didn't think I needed one. I had an approximate heat setting (based on charge amount), but if I was off, and the BT climbed too slow or fast, a slight tweak of a knob would get the curve right back on target. I could literally steer a roast profile to anything I wanted by turning that dial up and down. Very good for creating pretty roast curves; very bad for roasting good coffee. I found that my best roasts resulted from the fewest temperature changes.

When I moved to a gas drum roaster, I lost that handy little knob. Sure, I had a gas valve, but with the heat momentum of a drum, power changes happened much more slowly. If a BT was heading off in the wrong direction, there was nothing good that would happen from trying to correct it mid-roast. I was forced to have to plan ahead and execute that plan as flawlessly as possible. I've found most of the work in roasting coffee is done in the prep phase, before the coffee is even loaded in the hopper.

I call it "roasting hygiene". It starts with doing the same things the same way every time:
Warm up the roaster exactly the same way every roast session.
Weigh out the coffee to the gram.
Know what your charge temp will be,
what the temp at turning point will be and at what time,
what the time at dry will be,
what the time of 1C will be,
what the final temp should be and at what point in post 1C development, and
what the total roast time will be.
And while all that is going on,
what adjustments need to be made and at what times in order to produce a steadily declining RoR curve.
Whew! Who said roasting coffee well was easy?

Artisan, set up properly, gives me all the information I need to formulate a roasting plan for every coffee. It takes a little trial and error, but I can usually dial in a plan with a few roasts.

It starts with the first roast of a new coffee. In this case it's a new Kenya. Here's the very first roast from last night. I just so happens to look a lot like yours.



I brewed some this afternoon and even pulled a single shot. I did not get any baked or roasty notes, but it was also not dynamic and sweet. My first impression was it was roasted a bit too long. I think if I dropped halfway through that ending plateau, it would have been juicier and sweeter.

Some things to note: Very few temp and air changes. Normally I keep the air constant, but I was trying an experiment. This is also a 6# roast and I'm used to 8# charges. So for this roast, I was playing a little. But it still tells me what I need to add to the plan for the next roast.

For the next 6# roast I'd like the TP temp to be lower, so I will charger lower @ 380*. I will keep the same power settings, off at charge, 25% at 200*. I'm OK with 5:00 DE and this should get me there.

I started the roast with my damper at 25%. It's usually at 75% for the entire roast.

At DE I dropped the heat to 17%, my normal process. I turn down the heat when the coffee loses most of its free water. I tend to notice an ∆ET spike there, so I know BT is soon to follow. BTW, you need to get your ∆ET curve in the picture.

But this time, since I also opened the damper to 100%, I got a huge ET spike because the drum was now able to suck all the heat that was built up under the drum through the bean mass. This ran out after a minute, but it was enough to set the BT off on an undesirable plateau. This excess energy is what caused the crash. In order to stop the crash (and flick), I need to reduce the plateau to straighten the RoR plot like the blue line here:



At about 6:00 I closed the damper to 50%, which sent the ∆ET downward, even before I reduced heat to 8.3%. That made the crash worse, not better.

Then I opened the damper to 100% again, and the ∆ET spiked again. This likely reduced the crash, but added enough energy to begin the flick. So even though I turned the gas off at 8:25, the RoR did not go down. The result in the cup was that overdone taste.

So the next time I roast this coffee I will make the following changes:

Show this profile in the background. Artisan 2.0 now shows the setting values along with the time in background profiles (thanks Marko!)
Use the same 6# charge.
Charge at 380*
Set the damper to 75% and leave it there.
Drop to 17% 10s sooner.

And then see what happens.

Rytopa
Posts: 228
Joined: 7 years ago

#13: Post by Rytopa »

Rush wrote:Thanks for the input. I tried to find a profile that fit your description, I have tried all sorts of ideas in the past. This one is pretty close, it is 22.6% of rated maximum charge mass. The result is pretty good, probably least defective tasting. Still a hint of burnt/toast, but not grass like a roast with lower finish temp.
<image>
Is there a particular reason for adjusting the dampers after 1st FC?

I am asking because i am using a Huky roaster, which behaves more like a fluidbed/drum hybrid due to its thin drum walls and oversized fans. What i have found most helpful is to determine the airflow using the lighter trick and keep the airflow setting constant thru out the roast. This will help you gauge overall heat momentum of the roaster. Airflow changes can affect the dynamics of the roaster very dramatically, if you have trouble managing flicks and crashes, the last thing you would want to introduce is airflow changes to mange the ROR curve. Use heat adjustments to manage the curve and plan ahead.

Rush (original poster)
Posts: 90
Joined: 9 years ago

#14: Post by Rush (original poster) »

Rytopa wrote:Is there a particular reason for adjusting the dampers after 1st FC?

I am asking because i am using a Huky roaster, which behaves more like a fluidbed/drum hybrid due to its thin drum walls and oversized fans. What i have found most helpful is to determine the airflow using the lighter trick and keep the airflow setting constant thru out the roast. This will help you gauge overall heat momentum of the roaster. Airflow changes can affect the dynamics of the roaster very dramatically, if you have trouble managing flicks and crashes, the last thing you would want to introduce is airflow changes to mange the ROR curve. Use heat adjustments to manage the curve and plan ahead.
Without extra airflow the flick is crazy. I can get back up to 30 ror easily, with the heating elements off for over a minute. I'll test this for y'all.

Alan, thanks for the help!

@Almico Alan and everyone else, let me do some test roasts to see what happens. I haven't done a constant airflow roast in a long while, maybe I was missing something back then. I'll brown some pounds and post the results. Thanks for the help!

false1001
Posts: 279
Joined: 6 years ago

#15: Post by false1001 »

Rush wrote:Thanks for the input. I tried to find a profile that fit your description, I have tried all sorts of ideas in the past. This one is pretty close, it is 22.6% of rated maximum charge mass. The result is pretty good, probably least defective tasting. Still a hint of burnt/toast, but not grass like a roast with lower finish temp.
<image>
That profile looks pretty good to me, I'd just drop 30 seconds earlier. If you want to keep those DTR ratios maybe start your airflow changes a minute sooner and make them more gradual.

Rush (original poster)
Posts: 90
Joined: 9 years ago

#16: Post by Rush (original poster) »

Almico wrote:Rush, I understand the desire to "upgrade" equipment, but I'm not convinced it's the Cafemino here.
Alan, for a number of reasons beyond roast quality I would much prefer the air roaster to keeping my electric drum roaster, or buying a gas drum roaster. I know you think the more stable nature of the drum thermodynamics are better, but do you think it is possible to achieve the same roast quality (whatever that means) with the air roaster? Incentives include not burning my house down, not having to roast in 0°F weather, keeping the roaster and beans at a stable temperature all year long, etc. These are related to where I can install each type of roaster.

I have had some success roasting with this roaster for sure. It has proven hard to prevent crash entirely, but I have managed to get there occasionally. If I can even get to that point I would be happy with the switch. I promise to make as few heat adjustments as possible in any possible scenario, thank you for the good advice! :D

crunchybean
Posts: 463
Joined: 7 years ago

#17: Post by crunchybean »

Wow, this level of denial is epic. All three types of roasting machines have an affinity to roast to certain characteristics but all can be handled to roast like each other. There are famous roasters who use Deidrichs, Lorings and Probats (to generalize) and have figured out how to mitigate the C/F. The problem with mitigating the crash and flick of any roaster is up to you, NOT the machine. If you can't figure it out with your machine now, it's not going to be easier with any other, IR or otherwise. Across the site there are bits of info or all over the internet, go pay Rao a visit. Essentially, crashing is water loss too fast (flash heating and cooling, making your BT wave, flick is compensating during the heat wave and now your beans are skyrocketing. Loose heat more evenly before 1C or fly through the storm and change directions using, fan speed, heat or drum speed.) But the answer to your problem is a lack of reading/research/experimenting and not the machine.

From my own observations: air roasts do well with clarity of flavor, I have not tasted stronger/clearer flavors than on an air roaster. IR's at its best give flavors but never fail to have that warm "coffee" aroma. Drums give that full buttery/creamy coffee aroma.

All roasters can burn down your house and all should have proper ventilation.

Invest in less expensive beans so you can practice mitigating crash and flick. Reroast failed batches to dark roast and donate or save for breaking in a new grinder.

User avatar
Almico
Posts: 3612
Joined: 10 years ago

#18: Post by Almico »

Rush wrote:I know you think the more stable nature of the drum thermodynamics are better, but do you think it is possible to achieve the same roast quality (whatever that means) with the air roaster?
Yes, but it's much harder. With my drum, I can look away for 2 minutes post TP and for a minute or so before and after 1C. With an air roaster like the Artisan, you need to stay on it every second. It's OK for home roasting, but it would drive me crazy doing several back to back roasts.

Because of the nature of the design, the roasting air is changing over continuously. Any change in the air coming in will change your roast. The heaters are not constant on. They are PID driven and pulsed. I could never find out from Ken what the frequency was. I just knew my ∆ET plot was spiking up and down. Any fluctuation in supply voltage will affect the roast.

Rush (original poster)
Posts: 90
Joined: 9 years ago

#19: Post by Rush (original poster) »

crunchybean wrote:Wow, this level of denial is epic. All three types of roasting machines have an affinity to roast to certain characteristics but all can be handled to roast like each other. There are famous roasters who use Deidrichs, Lorings and Probats (to generalize) and have figured out how to mitigate the C/F. The problem with mitigating the crash and flick of any roaster is up to you, NOT the machine. If you can't figure it out with your machine now, it's not going to be easier with any other, IR or otherwise. Across the site there are bits of info or all over the internet, go pay Rao a visit. Essentially, crashing is water loss too fast (flash heating and cooling, making your BT wave, flick is compensating during the heat wave and now your beans are skyrocketing. Loose heat more evenly before 1C or fly through the storm and change directions using, fan speed, heat or drum speed.) But the answer to your problem is a lack of reading/research/experimenting and not the machine.

From my own observations: air roasts do well with clarity of flavor, I have not tasted stronger/clearer flavors than on an air roaster. IR's at its best give flavors but never fail to have that warm "coffee" aroma. Drums give that full buttery/creamy coffee aroma.

All roasters can burn down your house and all should have proper ventilation.

Invest in less expensive beans so you can practice mitigating crash and flick. Reroast failed batches to dark roast and donate or save for breaking in a new grinder.
After consulting with me, Mr. Rao strongly suggested I buy a new roaster and refused to work with me any longer. How does that make you feel about your comments? Additionally the crash and flick isn't even the reason for the desire to switch. All roasters have to be ventilated somehow, but not all roasters have to be ventilated in the same way. I think you might be trying to help, but are you doing so genuinely or just trying to mock me somehow?

Rush (original poster)
Posts: 90
Joined: 9 years ago

#20: Post by Rush (original poster) »

Almico wrote:Yes, but it's much harder. With my drum, I can look away for 2 minutes post TP and for a minute or so before and after 1C. With an air roaster like the Artisan, you need to stay on it every second. It's OK for home roasting, but it would drive me crazy doing several back to back roasts.

Because of the nature of the design, the roasting air is changing over continuously. Any change in the air coming in will change your roast. The heaters are not constant on. They are PID driven and pulsed. I could never find out from Ken what the frequency was. I just knew my ∆ET plot was spiking up and down. Any fluctuation in supply voltage will affect the roast.
This is excellent feedback, thank you so much! That is such a good point about the switch frequency, I hadn't even thought about it because my electric roaster has so much thermal mass that it is not detectable. In effect the heavy drum roaster structure and parts are like a huge inductor taking out the high frequencies, but on an air roaster the effect would certainly be seen by the beans and the probe!

Incidentally that is why my roaster is somewhat difficult to steer BT RoR, the heat chamber and surrounding metal parts take so much of the energy from the elements and then releases the energy at some point later. So the time constant on the roaster is extremely long. After the final roast my roaster takes over an hour to get within 10°C of room temperature, with an additional ventilation fan running in line with the exhaust for a good bit of extra CFM.

Just as an FYI, the switching period was likely 1s. So at 50% heat setting the SSR would be conducting for 0.5s and then off for the same.