Raost

[benjaminfleon]

One issue with this bean is that it seems to enter FC later than most beans (like 214-216). I'm guessing I should probably be modifying profiles to work around a higher FC temp?

Thanks for trying all these profiles! Yes, I definitely recommend adjusting the profiles so that the power profiling and constant fan speed match the time around FC. This is specifically the idea, as to avoid crashes before ~12% DTR, which is also where I usually adjust the negative boost to allow the PID to stick to the curve.

From visuals alone, it doesn't look like the newer versions have done a better job, but will see how they cup!

One question I have is that the charge temp seems a little low, especially with the Light Raost and Slow Raost. What's your thinking in lowering the charge temp so much?

I'd like to point out that, even though the newer versions might not look like they achieve better results (although I'm confident they do when you adjust them for FC), they are doing so in a much more flexible way by not relying as much on power profiling. One specific coffee might do great with power profiling throughout most of the roast, but that same profile won't be as easy to adjust to different coffees. Basically I'm trying to let the PID do all the heavy work, and just adjusting the timings for FC and DTR.

As for the preheat power, I think that has a lot to do with the coffee I've been roasting lately. I mentioned in another post that I've been roasting coffee that's super heavy on the processing, and it heats up super fast, so I adjusted accordingly. I'm trying to stay as close as I can to the profile line, so I also recommend adjusting the preheat (by 50-100W increments would be a good way to go).

This is very interesting! In the Kaffelogic profiles seems "L" to be very much coupled to end temperature (medium L=3.0 often 226 deg C end temperature). I still want a lighter roast to be well developed and maybe DTR can be to a help in that case. I haven't tried the Raost profiles yet but I will when it's time for next roasting session.

Hey! Yeah, as far as I'm concerned, the levels are given by the intersection of the profile with the ratio of end temperature and time —which you can adjust in the roast end by time ratio setting—, and that you can see as the red dotted threshold line.

If you're interested in a lighter more developed roast, I'd recommend a longer roast (maybe the slow or light versions). We often dismiss time as a factor for development, and the KL pretty much lets us try any profile (by which I mean, the shape of the curve and the relative proportions of each phase) stretched out in time.

Although color and end temperature seem to be the best current indicators of roast level, I'm still very interested in the area under the curve as a measurement. I've been recording AUC in my roasting log, as well as that number divided by the batch size (grams), so it's proportional, sort of as a relative area under the curve or something. So, for instance, I've found that when this number is under 10, it's pretty lighty developed, whereas when it's over something like 17, it's over-developed (FOR MY PREFERENCE, you may enjoy a much darker roast than me, and extraction plays a big role here, as well as water).

You should be able to roast and develop properly at almost any reasonable end temperature and color, but I think this indicator —AUC/Batch— is a more accurate representation of development throughout the roast.

[mr. bean]

Hey Benjamin, I've never paid much attention to AUC before. Can I confirm that you just type in the finishing time of the roast for 'end' and tick the box for the 'base' field?
I'm finding the Raost profiles are well under 10 when I calculate AUC/batch(g). For example the Raost v4 at 1.6 for the Ethiopian Yirg I shared above is 915.7 / 120g = 7.63
Even the FireStarter profile at 3.2 only gave me 13.
Where did you get the idea of dividing the AUC by the batch size? It's obviously making a significant difference that I've dividing by 120, but you're typically using 100g. But I don't think the Nano needs me to roast coffee for longer or hotter just because I'm using 120g rather than 100g... does it?

[benjaminfleon]

Hey Benjamin, I've never paid much attention to AUC before. Can I confirm that you just type in the finishing time of the roast for 'end' and tick the box for the 'base' field?
I'm finding the Raost profiles are well under 10 when I calculate AUC/batch(g). For example the Raost v4 at 1.6 for the Ethiopian Yirg I shared above is 915.7 / 120g = 7.63
Even the FireStarter profile at 3.2 only gave me 13.

Hey! Yeah, I basically type the finishing time and use the base temperature from the log. Now, I want to clarify that, because that measurement would be batch size dependant, it would be most relevant as a comparison between roasts of the same batch size. I'm thinking of asking a couple of friends who know more about mathematical modelling to come up with some sort of index that could be used to compare roasts of different batch sizes. Maybe incorporating preheat power or something. In any case, keep in mind that the AUC is an approximation of the energy transferred to the seeds, but there is also x amount of energy that is being lost to the environment.

Where did you get the idea of dividing the AUC by the batch size? It's obviously making a significant difference that I've dividing by 120, but you're typically using 100g. But I don't think the Nano needs me to roast coffee for longer or hotter just because I'm using 120g rather than 100g... does it?

I've been trying to be very thorough with my roast data lately, and including that measure just made sense to me. A while ago I listened to this podcast where Rao was a guest, and he mentioned that when he came up with DTR, he thought that it would make sense that there's some proportional relation between the time after first crack and total roast time. So, instead of just saying "two minutes after crack" as most roasters did at the time, he would say "20% DTR" on a 10 minute roast. I thought that was interesting and that maybe AUC could be adjusted in an analogous way, and batch size seemed like the most obvious way to go about it.

Like, theoretically the PID does what it needs to do in order to keep x amount of coffee on the profile, and you should adjust the preheat if you're using more or less coffee; but I suspect there must still be an inverse relationship between batch size and this broad idea of "roast development". So, with a given AUC, a smaller batch size would be more "developed".

To try to answer your question, I don't think you need to roast hotter or longer necessarily, but there should be a difference between a 100 gram and a 120 gram batch with the same curve and AUC/batch, and I would expect the smaller batch to be more "developed".

If I wanted to arrive at the exact same development with different batch sizes, I'd use time as an adjustment and use AUC/batch as an objective (I recently learned that you can adjust that by using the transform profile tool, although keep in mind you need to do it for both the profile and the fan speed, and adjust the power profiling as well).

In any case, you might not need or want more development; so it really depends if you're happy with your roasts!

[nrdlnd]

Hi Benjamin!
I have tried a couple of your profiles (two different roast with Raost V4 and one roast SlowRaost). I do usually taste them as V60 first and the Slow Raost came out best for me. I made an espresso from this roast also and it came out rather nice with a balanced acidity and sweetness. Maybe a little too tart. This is from only one espresso shot and I have not experimented with the settings. The bean was a washed Ethiopian Sidamo Wendo. The batch was 95g. Weight loss 18%. Light/Medium Light roast degree. DTR=24%. End temperature 219,3 and L=1.6.

210710_SidamoWendo_SlowRaostv2_1.6.png (54.7 KiB) Viewed 8847 times

Edit: I think preheat could be set higher. It's as default set at 675W and I think it could be higher especially with a hard washed bean from a high altitude as this Sidamo. I think it was a tendency to underdevelopment. I will set it higher next time. I think it's important to give more heat in the beginning for this kind of bean.

[benjaminfleon]

Hey, sorry couldn't reply sooner. I'm glad you're enjoying the coffee you get with these profiles! Yeah, I'd definitely adjust the preheat for your batch size and bean moisture/density. I'm currently roasting 80 gram batches most of the time, so you will need to adjust the preheat in order to have similar results with a larger batch.

Having said that, I think the easiest way to avoid underdevelopment with this profile (if you don't want to roast longer or get to a higher end temp) would be to drop your batch size. Let me know if you try 80 gram batches.

Along with that, I'm uploading the latest version of D-Raost (v3), which I've been using pretty often and has given me great results. I'm also uploading a slightly faster version (around 7 min to 219,9ºC), which is similar to the length of the Raost profile. I'm using these for 80 gram batches as well, and the beginning of the curve is very close to the profile with most coffees I've tried recently (1000-1200 mts naturals to 1800-2100 mts washed).

D-Raost v3.kpro

(2.1 KiB) Downloaded 966 times

D-Raost v3.png (109.54 KiB) Viewed 8812 times

D-Raost v4 fan.png (37.52 KiB) Viewed 8812 times

D-Raost v3 log.png (133.13 KiB) Viewed 8812 times

7 Min D-Raost.kpro

(2.12 KiB) Downloaded 1264 times

7 Min D-Raost.png (109.61 KiB) Viewed 8812 times

7 Min D-Raost fan.png (40.31 KiB) Viewed 8812 times

7 Min D-Raost log.png (122.09 KiB) Viewed 8812 times

If you want to roast more coffee with these profiles I'd recommend doing one of two approaches. First one would be increasing the preheat power (with these profiles, I'd say start by increasing 50W for each additional 5 grams or so). The easiest approach would be to drop your batch, but this would be close second.

The second thing you could do, which involves a lot more work, would be to adjust the length of the profile with the transform tool (in the tools menu) with a specific AUC/batch value. For instance: the AUC/batch value on the 7 min D-Raost for an 80 gram batch is 12.175 (974 / 80). So if I wanted to use the same profile, but, say, with 100 grams, I could adjust the time approximately by a x1.255 to get a similar AUC/batch value —I got this number by just trying out different values and getting as close as I could, took me like 5 minutes—. So, a x1.255 adjusted curve of that profile would have a AUC/batch value of 12.189 (1218,9 / 100), which would be close enough in my opinion.

Keep in mind that if you do this, you have to do the same transformation to the fan profile, as well as adjust all the power profiling and boosts to the new timings (which is not as annoying as I thought), and preheat.

I'm still figuring out the limits of how useful AUC/batch is tho, so don't take my word for it. I'm not saying that if you do this you will get an exact formula for adjusting the profile to different batch sizes, but I do think it's a step up from just messing with the preheat power.

Let me know if you try any of this!

B.

[nrdlnd]

Thank you for your great work!
I want to ask about the log D-Raost v3 log if you have a tendency to a ROR crash around first crack? I try to interpret these curves. What do you think?

[benjaminfleon]

I want to ask about the log D-Raost v3 log if you have a tendency to a ROR crash around first crack? I try to interpret these curves. What do you think?

No problem! With that specific roast there was indeed a slight crash, but not as bad as would have me concerned about super apparent baking.

I have found that with the D-Raost in general, it's very easy for the profile post FC to go very exothermic (if voltage isn't an issue), and keeping constant power during FC helps avoid that. At the same time, I've noticed that when there isn't a power profiling zone of 0 during FC, everything is way more random after FC. I think the key to make it work is to nail the timing with each coffee's FC temperatures.

For that log, for example, I might move the power profiling zone around FC a couple of seconds later, as I don't think it needs that big of an adjustment.

Let me know if you have any more questions!

B.

[rokkuran]

Hi, I've just got my 7E and have spent the weekend roasting with a few different profiles including Raost v4 and slow Raost v2. I'm looking forward to cupping them soon and pulling some shots.

Just wondering how long you normally rest beans you've roasted with these profiles? I was thinking of testing them at 3, 5, 7 and 14 days.

[benjaminfleon]

Hey Rokkuran!

Yeah, I usually cup the coffee the day after roasting and around 5 to 7 days later. I usually cup within those days in order to have an idea of flavors descriptors, taste balance and tactile, which I'll use for my customers. I've found the best flavors are after 14 days from roasting tho.

I reckon your approach would work well too!

Let me know how it goes, and thank you for trying my profiles!

[rokkuran]

So I blind cupped samples of some three day old samples of Yirgacheffe and Sidamo Ardi using Raost v4, Slow Raost V2 and Fast Raost v2 today.

I could easily tell the difference with the two Fast Raosts, which immediately tasted like a classic lighter filter style roast in the cup. I thought I had been able to nail down the difference between v4 and the slow but I got 3/4 wrong! I was however able to pick the differences in the two types of bean across those profiles.

I also made long blacks of the v4 and slow Raost, and I think as it stands, Raost v4 is beating the slower profile when making espresso based drinks, with a more pleasant body and after taste to me.