N3Roaster wrote:The darker you go, the harder it's going to be to apply both and get good results and if I find that I need to stretch yellow for a particular coffee, I usually also want to make an adjustment elsewhere to preserve acidity even if I'm intending a lighter roast. It's possible to do both under carefully chosen constraints and that can work with the right coffee, but it wouldn't be the first plan I'd reach for. Personally, I find that if you're trying to stretch out yellow it's better to plan based on a relaxed cubic spline, setting knots at the start of yellow, brown, first crack, 2nd crack if applicable, and end. That's going to minimize the curviness, you won't have crashes or flicks in the plan, and those sorts of plans are super easy to replicate with few and small heat adjustments once you've got a handle on anticipating what the machine wants to do.

Can you please explain the last part? i didnt get it

Thanks a head!

Sure, I did a blog post a few years ago. There are solvers online (I should really wrap a UI around the one I wrote and make that usable by others some day, but it's not a priority for me). Since writing that I've gotten better models for turnaround so I use those, but that's going to be highly machine-specific and I've gotten better at replicating this kind of plan than what's seen in the screenshots (getting rid of the saw-tooths in the air temperature profile).

Here are three roasts I did in the past week. The first two beans were Ethiopia Shakiso, a Royal CJ I've had sitting around for a while. I don't have the bean stats available.

I'm on a North 1 kg gas, and the charge weight was about 600 g in all cases. My usual plan is a 90s soak, followed by a burst of gas that declines. I may do a gas dip just prior to FC to lessen the pre-FC plateau/rise, and hopefully the crash/flick. This bean tends to cooperate in that regard- it's not very crashy, and it tends to have a very high FC temp. Typically 10°F higher than most. I'm using 1.5 mm probes. Air is constant at 40%, but I raise it to eliminate volatiles and moisture into FC. My strategy through FC is to titrate gas to the ET, trying to maintain a steady to rising ET. I figure BT will do what it wants to do, but I can get most immediate feedback on changes in ET. Burner settings can be interpreted by moving a decimal point one spot to the left for values over 1.0 (ie B12 = 1.2 kpa, but B6 = 0.6 kpa).

The first roast I charged on the hotter side with more gas (1.8 kpa). For some reason the phases didn't appear on the graph. As you can see, there's a bump prior to a crash, even with a gas dip. This coffee came out kind of sour and cereal tasting. Not good. I tend to get these kinds of effects in short, hot roasts.



The second roast I charged lower, with less gas (1.6 kpa). TP was lower, confirming (to me) less energy in the drum. FC came later by about 1 min, and there is less of a crash. It also finished hotter, and was a considerably longer roast overall. This coffee came out less sour, no cereal notes, but some nice strawberry fruit flavors.

As a theme I've noticed lower charge temps, less crash, and longer roasts that finish at higher temps tend to taste better to me. I'd love to get to a point where I can roast beans light that are still developed and flavorful. I think one of my unaddressed issues is that I need to experiment with higher charge weights, closer to 1 kg. I struggle on this roaster to stretch drying out to 5 min without using very low charge temps, like 300-320°F. My strategy also tends to fall apart with crashy beans, like Kenyans. I'd like to improve on that if you have any suggestions.



For a cross-bean comparison, I ran a similar strategy with a Colombian CJ (that's been sitting around for a very long while!). As you can see there's more of a crash. I haven't tasted it yet (we're 3 days out now).

Almico wrote:The only way I have found to stop it, is to reduce the heat more than usual approaching 1C and then hitting with every one of the 78,000 BTUs I have available for almost a full minute.

So Scott's most recent Instagram post shows this exact method is not good.

For my roaster, I can completely eliminate a bad crash with air adjustments only. But IMO, air vs gas is very similar. Both result with a sudden increase of heat to prevent a drop in bean temps. I've tried both with no significant changes of flavor. Much more significant to flavor is the ET going into FC. Too high and you get that "shelf" and "flick" after FC.

Perhaps I still haven't experienced the results of the perfect ROR curve. But after years of roasts and actually accomplishing many perfect Rao curves and not so perfect for the same beans, I think I have the experience of both. And for me, the advice from Neil is a very good one. Time is much better spent on experimenting with different profiles and shapes.

EddyQ wrote:So Scott's most recent Instagram post shows this exact method is not good.

For my roaster, I can completely eliminate a bad crash with air adjustments only. But IMO, air vs gas is very similar.

My plan with this coffee is definitely not Scott approved. But it works well. I have not found any other way of stopping the crash.

Air and heat are definitely not the same thing. Scott's advice is to leave the air alone. Find a setting that evacuates most of the chaff and leave it there. Then learn to control the crash/flick with heat only. If you are not doing it this way, you have likely not yet experienced its benefits. It's not a subtle thing.

Almico wrote:Find a setting that evacuates most of the chaff and leave it there. Then learn to control the crash/flick with heat only. If you are not doing it this way, you have likely not yet experienced its benefits. It's not a subtle thing.

I've controlled the crash with gas only. No difference. I've even controlled the crash with only heat, no changes to air or gas. But the latter will not work with a crashy bean used for a darker roast unless you like baked coffee.

One of these days, we need to ship each other some coffee and do some blind tasting. Match the profile blindly to the coffee.

Well when i had rao curves i had little success.

had better with more elastic ones, this particular batch didnt tast so good but other with similar idea was good.
The plan is basically is gas off for 1:30 , then up the gas , then down when approaching yellowing to stretch as suggested to sweetness, then go a little bit faster at browning to approach first crack momentum and reduce the gas gradually after 1C, stop just a little before second crack

will like your opinion, thanks!

Seems like your charge temp is too low, there is not enough heat momentum, try to adjust ur heat and aim towards going in 1fc at around 30% percent heat.

ymg wrote:Well when i had rao curves i had little success.

had better with more elastic ones, this particular batch didnt taste so good but other with similar idea was good.
The plan is basically is gas off for 1:30 , then up the gas , then down when approaching yellowing to stretch as suggested to sweetness, then go a little bit faster at browning to approach first crack momentum and reduce the gas gradually after 1C, stop just a little before second crack

will like your opinion, thanks!

image

Anyone try the latest RDK?



I found the differences a lot less pronounced than the kit from Aus roaster Criteria.
Cupping Roast Development Samples

Still, in the Rao / Regalia Good there is a distinct difference in the upfront sweetness, vibrancy and structure of the brew from hot to cold.
Seems like a very high quality Colom green the baked is still drinkable im guessing many people might even like the flattening / homogeneous flavour.