Kaffelogic for the USA is on Indiegogo - Page 6

#51: Post by **ducats** » June 30th, 2022, 8:09 am

lukehk wrote:I've come to settle on the Raost profiles for most of my roasting. I've not used DRaost. I think it's an adaptation of the original Droast (profile from a different person) which is a popular profile that is well liked. Raost V4 is my go-to and has worked well for a variety of beans. I reduce the preheat for delicate beans and finish earlier. Anything from 213deg finish for a geisha to 219 for others. I managed to get in on the daterra masterpiece split here on HB and that was really good finishing at 215-7 depending on the balance I wanted between black tea and citrus. Ninety plus Kambera was also very good finishing 215-217. There is also a long slow version of raost as well which I use occasionally if the need arises. The creater of the profiles talks about his theory of a constant fan speed at FC as well as constant power zones. I should start adjusting fan speed more specifically for FC to test this but it's been good without this level of tinkering and I've been playing around a lot with rehydration so one thing at a time. I would say with raost that I have found it follows the profile better without the flat power zones....... Whether it tastes better or worse I don't know as I have not compared.

Good to know. How long have you been using KL? I see Raost v4 in "about this file" recommends 1.7-1.9 for roast degree but going to 215C according to the profile is 1.0 and about 51 sec after 1C for about 6C increase. No underdevelopment for those? I'd guess not if you're doing it

What's your go-to brewer/recipe for filter and what batch size are you using for Raost v4?

#52: Post by **lukehk** » June 30th, 2022, 12:57 pm

I've been using it a couple of years, about 350 roasts. My batch sizes are usually 80g as this is what the raosts were designed on. He has also posted profiles for many ninety plus coffees so I think smaller batches were chosen to make those expensive coffees go further. Some also work well at 100g too. When rehydrating it's about 75g+5g of water. For some coffee I don't need to adjust the profile for others I need to increase the fan speed at the beginning to get the beans moving. The profile for the Kambera is basically V4 stopping at 215. Maybe some other tweaks with fan or power zones but essentially the same curve. I used the same for a Panama Esmeralda Private Collection Geisha Natural but stopped it even earlier. I don't imagine(?) It's different to other roasters in that delicate or heavily processed coffees need less development time and less less initial heat. As the PID adjusts the power to ensure temperatures follow the profile the profile is quite flexible and it depends how much development you like. If a coffee does not do so well I chose the slow Raost version. The profile for drima zede is particularly long..... Fishing at the same end temp as the others. The profile is a starting point and he says he adapts it for each bean but I think you get a good initial roast (although I remove the power zone usually). The ability to manipulate any aspect of the profile curve, fan curve and specific zones is excellent. When you mark FC will also affect the figures. Patrick Rolph from April coffee says mark at the very first crack others including myself wait for a few clear consecutive cracks. The temp of first crack also changes between beans and level of moisture. I've had good results for anything from 45 seconds after FC to 1 minute 30 but it's usually around 1 min and first crack has never finished. In saying that when rehydrated to 16-17% I have found I can get through to the end of FC in around minute...... Which I can not do without it. The levels are user set and not comparable across profiles so I tend toward using end temp.
My go-to brew method is Jonathan Ganges 10 minute aeropress or a kono dripper 4x60/65ml to 15g coffee. Grinding on a Bentwood at 500m with a water mixture of 150ml evian to 850 distilled water

#53: Post by **mtbizzle** » July 1st, 2022, 4:13 am

thanks for sharing!
one idea from the Hoos video that seemed like it could be reasonable to try on the KL - make 'bigger' fan speed adjustments during yellowing, perhaps near the end of the yellowing phase. This seems to align with practice that Royal uses for Ikawa. Some other possible reasons in favor of this-
-fan speed needs to be highest at the beginning
-I believe general consensus is, yellowing might be the least 'important' and sensitive phase of the roast. any issues brought up by changes in fan speed might have the least impact here.

perhaps this could leave only subtle fan speed adjustments after the end of drying.

Love the idea of looking at Ikawa profiles - e.g. those developed by professionals, community favorites - for ideas to try re. fan speed.
"Standard practice these days for Ikawa roasting is to put our Crown Jewels through a gamut of 4 reliable profiles at what we'd consider to be "drinkable" sample roast levels. "

From Royal. it looks like 3 of the 4 profiles are "crown standard SR 1.0", "crown maillard +30 SR 1.0", "crown inlet sample roast 2022". You can see them here, and download an excel file that has the fan settings. https://royalcoffee.com/product/3427097000013083031/

I would also note that folks at Royal are explicitly happy to have an ugly curve if it tastes better... See the blue and red curves in the link above, the temp probe reading flatlines at crack. I looked at one such roast in detail and the temp reading was 211-213 for a whole 90 seconds.

I think there's almost always an aesthetic element to creating roast curves as well... a desire to round off the edges, make something look more like what we believe to be true about stalling, exotherming, or heat deltas... I know my profiles look like they stall at the end. But after Color Tracking and cupping and going back to the drawing board time after time to get the right color and development time with a declining fan speed, I found that the results in the cup tend to override my aesthetic sensibilities.

July 5th, 2022, 5:02 pm

I leave on vacation and find that thread totally off topic (and hope one of you won that giveaway I missed to),

Can a moderator maybe split if and give the roasting discussion a better thread name? It has nothuing to do w/ the Kaffelogic topic, really.

July 5th, 2022, 10:32 pm

I've had a look at the All of the discussion about roasting seems to be geared towards the capability of the kaffelogic, from what I can see, which I think was what we thought your original post was asking about (at least in part). I can't really see how we can hive off any of the individual bits.

July 6th, 2022, 1:28 pm

I guess it's hard sometimes to tell whether the discussion (or a graph) is about the KL, the ROEST, or something else (few mentions of the Ikawa Pro too). At least for me.

How does one answer the questions: "Should I back this if I want 1/ to do small batch roasting for drinking (not sampling for buying) and 2/ learn about roasting with some playability in params etc.?" And since a lot of you have lots of experience in general... say you buy an Ikawa Pro and get beans and roast profiles from them, are those meant for drinking rather than cupping? Some articles (incl. that one from Rob Hoos that is mentioned in a recent post somewhere about the Bullet) make a clear distinction between roasting for sampling vs drinking.

#57: Post by **luca** » July 7th, 2022, 1:55 am

Frenchman wrote:I guess it's hard sometimes to tell whether the discussion (or a graph) is about the KL, the ROEST, or something else (few mentions of the Ikawa Pro too). At least for me.

Oh ok, sorry, let me see if I can explain the relevance a bit ...

The graphs may be from different machines, but they are to discuss and illlustrate points relevant to the KL. The KL has one probe, that sort of sits above, sometimes in, the circulating bean mass. The question is what on earth does that temperature reading mean, how do you use it, and how comparable is it to what you see from other roasters? (The same question can and should be asked for every roaster on the planet.) The other roasters have been discussed to try to answer these questions. I posted graphs from the roest because it has many temperature probes and it is also a roaster that has a high flow of hot air. If nothing else, those graphs show that what temperature information you get is very dependent on where the probes are positioned.

The reason why this is important is because it goes to how you use the machine and what sense you make from the graphs. The most widely accepted idea of what a roast graph should look like is what's sort of attributed to Scott Rao, which is basically that the bean temperature ought to follow a curve with a derivative (the "rate of rise", because for some reason coffee has to use this weird term instead of the highschool algebra term we should all be familiar with) that is a linear straight line smooth decline. Lots of people accept this uncritically. However, that is based on a bean temperature probe reading in a traditional commercial drum roaster. The question is whether the temperature reading in the kaffelogic is analagous and can be used in this way. Many people gloss over this issue and assume that it can. Some people are sceptical. Others believe that their experience is that you can apply this methodology. If you can't apply this way of working out what roast profile to make, then the question is what principles do you apply to construct a good roast profile.

How does one answer the questions: "Should I back this if I want 1/ to do small batch roasting for drinking (not sampling for buying) and 2/ learn about roasting with some playability in params etc.?"

There is information in this thread about the temperature data questions that I have posed above, and you can sort through it and come to your own conclusions. I don't have a definitive answer for you.

What I did ask ducats to test was if he could get a really fast roast out of the 110V KL or if power would be an issue, and it looks like his results were kind of promising that you can, at least for the dose he tried and the roast level he got to. This is promising because it means that if you want to roast slower, for the same dose, the machine should be able to, since it requires more heat and more power to roast faster. Whether the roaster has enough power to do that for a higher dose is a question.

We would all like things to be simple, but sometimes complexity exists in the world.

July 8th, 2022, 12:26 pm

luca wrote: We would all like things to be simple, but sometimes complexity exists in the world.

Very true, and something I deal with every day (and relish) in my job. But I am very qualified and experienced at my job, and not at all in coffee roasting! Thanks for explaining the relevance of the posts that I'll re-read, and in advance, for putting up with naive questions I'll post as I learn (like this and this).

#59: Post by **luca** » July 9th, 2022, 1:15 am

You're welcome and there are no stupid questions; feel free to ask anything. Some of the discussion above may have been a few of us sort of implicitly responding to things that we had each raised, so sorry if I made it difficult to follow.

July 10th, 2022, 6:25 am

I have had a KL for about 8 months now, the company is actually based in New Zealand. The units are popular in NZ and Australia. I use their dark and steady profile to roast an espresso blend which I use to make milk based drinks. I have a pretty standard mix of Brazil Washed, Brazil Honey, Indonesia an PNG beans (each 25%). It gives a pretty good full flavoured espresso.

https://kaffelogic.com/community/viewtopic.php?f=2&t=11

I am pretty happy with the KL unit. I am waiting for the new chaff collector, which should be released shortly, they will also release a expansion unit to increase the roasting size from 110gm to 200.

I was using the provided software for a while, but don't really use it now as I found the dark and steady profile worked pretty well. I am very happy with the KL, it is my first roaster so I am not able to compare it with others. I had some issues/questions, and the customer support was pretty good.

The community is pretty active there on the KL forums, I also found that helpful.