Trying to match small drum roast
-
kaffelogic
- Site Admin
- Posts: 514
- Joined: Thu 18 Oct, 2018 9:38 am
- x 74
- Contact:
Re: Trying to match small drum roast
Hi Steve, great post. Sorry I have been a little late chiming in, but here goes...
There is a temperature envelope feature in our Studio software that is intended to support what you are doing. The theory is that the temperature differences between one roaster and another can be summarised by capturing the temperatures at specific milestones in the roast. For example, if you know the temperatures of colour change, first crack, and second crack this should be sufficient for accurate temperature conversions, at very least it should be applicable for a range of similar profiles if not all profiles.
The idea is that you would do exactly what you are doing: establish a Kaffelogic profile that is equivalent to a drum roaster profile, and use what you have learned to set up a temperature envelope. This should allow you to import a profile from your drum roaster, make experimental changes to it on the Kaffelogic until you have made an improvement, and then export it back to the drum roaster. The improvement should then be manifest in the drum roast.
It's still a bit of an open question how well the temperature envelope will work on its own. It may be necessary to perform some time transformations as well (using the 'transform' tool in Studio) to allow for the greater heat capacity of a larger batch and for the greater delay in a typical drum roaster's temperature probe.
What you are doing is cutting edge research, so keep posting and I will provide any help that I can.
Chris
There is a temperature envelope feature in our Studio software that is intended to support what you are doing. The theory is that the temperature differences between one roaster and another can be summarised by capturing the temperatures at specific milestones in the roast. For example, if you know the temperatures of colour change, first crack, and second crack this should be sufficient for accurate temperature conversions, at very least it should be applicable for a range of similar profiles if not all profiles.
The idea is that you would do exactly what you are doing: establish a Kaffelogic profile that is equivalent to a drum roaster profile, and use what you have learned to set up a temperature envelope. This should allow you to import a profile from your drum roaster, make experimental changes to it on the Kaffelogic until you have made an improvement, and then export it back to the drum roaster. The improvement should then be manifest in the drum roast.
It's still a bit of an open question how well the temperature envelope will work on its own. It may be necessary to perform some time transformations as well (using the 'transform' tool in Studio) to allow for the greater heat capacity of a larger batch and for the greater delay in a typical drum roaster's temperature probe.
What you are doing is cutting edge research, so keep posting and I will provide any help that I can.
Chris
Re: Trying to match small drum roast
Thanks Chris, that is quite helpful and I was not aware of this feature. I will have play around with it at some stage when i feel like things are close enough.
Haha, im just tinkering way beyond what most people seem to think is sane. Your profile design studio, now thats cutting edge!
Haha, im just tinkering way beyond what most people seem to think is sane. Your profile design studio, now thats cutting edge!
Re: Trying to match small drum roast
UPDATE:
Very happy with this style of profile on the Nano, much sweeter and more balanced roasts (I will happily sacrifice some acidity for increased sweetness and body).
Trying to follow all of Scott Rao rules to the letter.
Aiming for close to equal parts / time for the 4 stages of the roast =
Start to 100BT (assume 110 on Nano) 25%
100 to full yellow 25%
full yellow to first crack start (FCS) 25%
FCS to finish 22 to 25% = med - med dark roast.
(not an exact science!!)
Yet to nail / get my small drum able to manoeuvre close to the Artisan profile, that picture is a screen shot of the "ideal", so still very much work in progress.
Very happy with this style of profile on the Nano, much sweeter and more balanced roasts (I will happily sacrifice some acidity for increased sweetness and body).
Trying to follow all of Scott Rao rules to the letter.
Aiming for close to equal parts / time for the 4 stages of the roast =
Start to 100BT (assume 110 on Nano) 25%
100 to full yellow 25%
full yellow to first crack start (FCS) 25%
FCS to finish 22 to 25% = med - med dark roast.
(not an exact science!!)
Yet to nail / get my small drum able to manoeuvre close to the Artisan profile, that picture is a screen shot of the "ideal", so still very much work in progress.
- CompareDrumNano.jpg (48.05 KiB) Viewed 5724 times
- NanoMed.jpg (42.51 KiB) Viewed 5724 times
- DrumArtisan.png (87.65 KiB) Viewed 5724 times
11minMED.kpro- (1.85 KiB) Downloaded 320 times
Re: Trying to match small drum roast
I have roasted Bolivian beans with Steve's 11MinMed profile while ago (2/Oct).
100G -> 82,4G , level 3.0 and 25.4% DTR
Brewed last night ( pour over) and it still very sweet and fresh taste, every bit was enjoyable.
So Thanks for sharing this!
PS. Along with light roast version of yours, both are great.
PPS I kept beans in foiled zip lock bag with one way bulb that made for roasted beans.
100G -> 82,4G , level 3.0 and 25.4% DTR
Brewed last night ( pour over) and it still very sweet and fresh taste, every bit was enjoyable.
So Thanks for sharing this!
PS. Along with light roast version of yours, both are great.
PPS I kept beans in foiled zip lock bag with one way bulb that made for roasted beans.
Ben
Re: Trying to match small drum roast
Excellent Ben!
I have been playing around with slight modifications of these profiles and getting even better results.
Mostly, just a bit straighter through the mid section and a much slower finish to a lower end temp. I have realised that I prefer this to shorter post crack times that finish at higher temp.
I will post some profile updates soon.
I have been playing around with slight modifications of these profiles and getting even better results.
Mostly, just a bit straighter through the mid section and a much slower finish to a lower end temp. I have realised that I prefer this to shorter post crack times that finish at higher temp.
I will post some profile updates soon.
Re: Trying to match small drum roast
Excellent thread. I'm going to give your 11MinMed profile a run today.....thanks for all your hard work Steve!
Cheers Mark
Cheers Mark
Re: Trying to match small drum roast
Thanks Mark.
Depending on what flavours you prefer in coffee? These may not be punchy enough for some people especially in milk. My aim to is develop the sweetest possible roast while maintaining the inherent characteristics the seed has to offer, with as little roasty flavours as possible. For those who like darker roasts I am still fine tuning a similar profile 12 min to start of second crack with a declining ROR / around 25% DTR or you can just play around with the curve so that it enters first crack a little quicker and have the ROR not drop off so quick for a higher end temp.
The following profiles are where I am at with matching roasts on my little drum roaster (400g). Actual usable batch size (without burning / underdevelopment) is 225g to 300g depending on how long I want to roast to for.
I have worked out the difference in my temp readings are definitely not linear and the roasts below are a pretty good match, except the drum roast needs a bit lower finish temp / -15sec.
The drum roast required some thinking outside of the common ideas around small vs large drum roasters box. Andy over at CS has always said that the higher charge temps do not apply to these tiny roasters and I think he is right.
Drum is 250g and the Nano is 100g.
IMPORTANT: With the ROR dropping so low toward the end of the roast on the KL / nano profile it usually requires that you end the roast manually, otherwise it can drag on (go into a bake before reaching your desired end temps OR end the roast early.
I am logging first crack at the very first pops I hear (I find this to be the most consistent).
End of first crack logged when I hear the last 1 or 2 pops.
Its also worth noting I have changed the " roast end by time ratio" in engineer level settings to 0.5 from the stock 0.33333 which I have found to help with this style of roast.
- Drumroast.jpg (25.94 KiB) Viewed 5672 times
- Drumroast.png (104.11 KiB) Viewed 5672 times
- Guat.jpg (30.31 KiB) Viewed 5672 times
- Compare.jpg (30.35 KiB) Viewed 5669 times
Guat.klog- (71.19 KiB) Downloaded 280 times
- Drumroast.klog
- (7.02 KiB) Downloaded 283 times
- Light-Med11min.kpro
- (1.84 KiB) Downloaded 280 times
Re: Trying to match small drum roast
Excellent .
Another curve to try, I have save it to Stave drum-ML in my library so I know where it comes from.
I may try Mexican or Colombian next time.
Good experiment backed with knowlage and experience is great to follow ! I have lot's to learn before making drastic change to y own experiment. I'm getting back and forth few variation now.
Another curve to try, I have save it to Stave drum-ML in my library so I know where it comes from.
I may try Mexican or Colombian next time.
Good experiment backed with knowlage and experience is great to follow ! I have lot's to learn before making drastic change to y own experiment. I'm getting back and forth few variation now.
Ben