Help with getting sweetness

[TheBean]

Many thanks for your advise - definitely going to try MASL profiles!

[terapico]

... best case scenario, combine a lower temp and shorter shot.

But, yeah, if you can fix the roast, even better. I truly enjoy the altitude based profiles that Wayne designed. For someone not yet confident to do their own profiles, they seem to work well 80-90pct of the time. I find bracketing the roast lervel corrects the remaining 10-20%, and I just keep a little log per bean so I can remember what to roast!

I have just switched to KL Nano from a Hottop. I went directly to MASL profiles, and the first batch of Huehuetenango showed marginally darker colour and (very marginally) lesser flavour compared with my last HT roast of that bean, a type I like. I will drop the level for the remainder of those beans, not having enough left to bracket, then get serious with some new other beans. I agree with your comments and am confident that a bit of profiling over automation is going to give me a good and more reliable result than I had with the HT through weather variables in a non-coastal climate.

[TheBean]

Hi @SniffCoffee, it’s been a while since you’ve started roasting coffee so you’re now a pro while I’m a newbie… I’m having the same “charred” aromas you had at the beginning- how did you manage to overcome it? I’ve tried different levels of roasting, from 2.1 to 4.2, but the “charred” aroma always there.

Purely my two cents, but in my experience the K-Logic Classic profile produces such notes at default and even in the 2.0-3.0 range.

I tend to use the other core profiles that go by MASL (metres above sea level) for the origin of the bean, stick to between 2.6 to 2.8, and come away with something nicely caramelised and sweet - balanced to my palate.

Another thing to try is how you brew a char-like roast. If possible, lower the PID temp. If not, flush, wait 10, 20 or 30 seconds, then brew. Either way you're waiting for a lower temp on boiler temp recovery.

And set your ratio to 1:1.5 or 1:1.25 - meaning, if 18 gms in, 1:1/5 is 27gms out, and 1:1.125 is 22.5gms out.

Cutting it short can remove those cruddy notes. And in best case scenario, combine a lower temp and shorter shot.

But, yeah, if you can fix the roast, even better. I truly enjoy the altitude based profiles that Wayne designed. For someone not yet confident to do their own profiles, they seem to work well 80-90pct of the time. I find bracketing the roast lervel corrects the remaining 10-20%, and I just keep a little log per bean so I can remember what to roast!

I’m new to the rotating - How do you “bracket the roast level” ?

[terapico]

…
I’m new to the rotating - How do you “bracket the roast level” ?

If the default roast level is, for example, 3.2 then try 3.0 and 3.4, forming a bracket around the normal roast level. If you prefer one of those results, go further in that direction until your taste buds tell you to come back a bit. The idea is to find end points for your preference while narrowing your search range to an area that suits your taste.

[Eldorado Green]

Hi @SniffCoffee, it’s been a while since you’ve started roasting coffee so you’re now a pro while I’m a newbie… I’m having the same “charred” aromas you had at the beginning- how did you manage to overcome it? I’ve tried different levels of roasting, from 2.1 to 4.2, but the “charred” aroma always there.

Purely my two cents, but in my experience the K-Logic Classic profile produces such notes at default and even in the 2.0-3.0 range.

I tend to use the other core profiles that go by MASL (metres above sea level) for the origin of the bean, stick to between 2.6 to 2.8, and come away with something nicely caramelised and sweet - balanced to my palate.

Another thing to try is how you brew a char-like roast. If possible, lower the PID temp. If not, flush, wait 10, 20 or 30 seconds, then brew. Either way you're waiting for a lower temp on boiler temp recovery.

And set your ratio to 1:1.5 or 1:1.25 - meaning, if 18 gms in, 1:1/5 is 27gms out, and 1:1.125 is 22.5gms out.

Cutting it short can remove those cruddy notes. And in best case scenario, combine a lower temp and shorter shot.

But, yeah, if you can fix the roast, even better. I truly enjoy the altitude based profiles that Wayne designed. For someone not yet confident to do their own profiles, they seem to work well 80-90pct of the time. I find bracketing the roast lervel corrects the remaining 10-20%, and I just keep a little log per bean so I can remember what to roast!

This is my experience as well - though I'm only starting out I'm really struggling to get repeatability and to avoid the charring and burnt taste... Despite quality coffees.
I'm finding that almost all of the profiles at around Level 3 are burning the coffee. Could this be about the difference in ambient temperature between Sydney and NZ? Or could my machine just be running hot?
The second problem I have is that when I drop the altitude profiles to 2.4 or 2.6, the DTR goes too long or two short...
I'm trying to stabilise everything this week by just going back to a single Bean - the Colombian Volcan galeras Supremo - and just roasting that... I guess there are so many variables in roasting. It's an adventure.

[Geronimo]

This is my experience as well - though I'm only starting out I'm really struggling to get repeatability and to avoid the charring and burnt taste... Despite quality coffees.
I'm finding that almost all of the profiles at around Level 3 are burning the coffee. Could this be about the difference in ambient temperature between Sydney and NZ? Or could my machine just be running hot?
The second problem I have is that when I drop the altitude profiles to 2.4 or 2.6, the DTR goes too long or two short...
I'm trying to stabilise everything this week by just going back to a single Bean - the Colombian Volcan galeras Supremo - and just roasting that... I guess there are so many variables in roasting. It's an adventure.

Greetings

As an aside are you getting adequate bean circulation. I’m wondering if it’s worth dong a fan calibration. The info is on this site somewhere.

I find generally 100g of beans is a good weight for me, and did have to crank the fan speed calibration up a bit.

On the stock profiles. I notice an undesirable ashy flavour if the level is set to the dark level range as per the roast profile notes. I’ve had to drop the levels quite a bit to reduce the ashyness.

Not sure if you are roasting different altitude beans together, but if you are, then that can cause under and over roasting. Although in saying that if you can get some unique flavours compared to post roast mixing of the beans.

Cheers
G

[Eldorado Green]

Thanks for that.
The machine is new, so not sure if it would need calibration - but I'll have a look around and see if I can do it.
But yes, I'm finding anything from around 2.8 up pretty dark / ashy, and have roasted a few using the dark profiles - for instance Darkside and Droast - and they were pretty much undrinkable.
I'm still learning, of course - but the one thing I've learned most quickly is that it's not just about picking a profile and pushing a button.
It's good to hear you have had difficulty too. Thanks.
Just out of interest, do you normally stop the profile yourself, or do you let it run until it stops? For instance, today I used the Ninja ii setting on some Columbian Galeras. I set the roast level to 2.4, but it was going way over, so I stopped it when it got to 24%DTR, because I didn't want another undrinkable batch... But then it told me the end roast level was 1.7... Hmmm.
I haven't tasted them yet...

[Geronimo]

Thanks for that.
The machine is new, so not sure if it would need calibration - but I'll have a look around and see if I can do it.
But yes, I'm finding anything from around 2.8 up pretty dark / ashy, and have roasted a few using the dark profiles - for instance Darkside and Droast - and they were pretty much undrinkable.
I'm still learning, of course - but the one thing I've learned most quickly is that it's not just about picking a profile and pushing a button.
It's good to hear you have had difficulty too. Thanks.
Just out of interest, do you normally stop the profile yourself, or do you let it run until it stops? For instance, today I used the Ninja ii setting on some Columbian Galeras. I set the roast level to 2.4, but it was going way over, so I stopped it when it got to 24%DTR, because I didn't want another undrinkable batch... But then it told me the end roast level was 1.7... Hmmm.
I haven't tasted them yet...

There have been a few new machines that have needed a fan speed increase from the default.

Just out of interest, when are you recording the start of first crack?

I run to the desired roast level that I have set initially, but I have a good record of what it needs to be. Initially I did stop it if it the development time was getting into silly numbers and beans were to dark. Typically I have found my development time is around 3 minutes plus or minus 15 seconds or so. Other people will obviously have different times, but that is on average mine.

I notice on the stock profiles the development time for darker roasts gets up over 30%, which isn’t necessarily wrong. Your heading up to the second crack region.

You don’t necessarily need to be between 20 to 25% development time ratio. Scott Rao who first mentioned this, said it is a guide, but not necessarily set in stone.

Scott Rao says, and I quote him here, “As the guy who invented DTR, I’m begging roasters, please do not drop batches based on DTR. Roast colour should be the determinant when to drop the batch. Bean temperature can be a useful proxy for colour”. Further on mentions “DTR can be useful for analysing the balance of a roast profile.”

He added a PS at end of blurb: “A seperate but related issue is that DTR is not a great indicator of roast level . . . for example, you may reach 20% DTR early in first crack, or you may reach at the very end of first crack. For my colleagues who think “20%DTR is dark,” please realise that my 20%DTR maybe lighter than your 16%DTR.”

Food for thought.

[fnq]

Hi Eldorado Green

As well as agreeing with Geronimo's good words above, i would suggest double checking your logs.

Are you getting a good read and logging FC? I have found this does vary by bean but if in doubt I factor about 210- 211 degrees as commencement of 3 or so quick cracks.. Some probes/ logs on here are closer to 206 degrees which is neither good nor bad but it does alter the DTR calcs.

The other point i may ask you to consider is whether or not your logs are sticking closely to your MASL profiles.. Stick some logs up if you would like to share.... see if you are flatlining power.
Geronimo's point about 'how each bean get through the development zone' -- time/ momentum / sharply rising / flattening/ what ROR does etc is important.

I have experienced ashy tones when sometimes roasting into the mid 4s and above,, but not down in the territory you are looking at on those profiles.

That is a good bean (Columb VG)for medium to medium dark roasts . I have used 1500-2000 MASL @ 3.2 to 3.4 and noted lovely choc tones at the finish. For testing purposes I found that this bean drinks quite well using the Altitude RTD profile ( so you don't have to wait 5 days or so to rest the beans.....

So your mission is clear..... try and log data and engage a 1500-2000RTD profile at say 2.4 ( forget about DTR atm)....drink 4 cups of coffee a day to enable the experiment to be repeated tomorrow with slight tweaking , if you dont make any progress then give The KL team an email , they will work through logs etc Good luck Darryl

[Eldorado Green]

Thanks everyone for your input and your thoughtful responses.
I've gone back and double checked for power flatlining - I think on a couple of the roasts that did burn there was some sort of interruption to the power or the KL just didn't get what it needed. I've also gone back to a simpler learning strategy - just using one or two varieties at a time and bracketing a little more carefully. The last few roasts have been better.
@geronimo - yes it's good to be reminded DTR is only a guide... and I've been looking closely through this forum and seeing that the DTR does vary, and modifying my expectations.
@fnq thank you also - -my last couple of roasts seem to have worked well - but tomorrow I will try that Colombian on the RTD profile you've suggested and will report back.

This is a lovely community - thanks all for weighing in on my anxieties about getting this right. It's early days, so I'm happy to still be learning.