topbanner.gif
Login
Username

Password




Not a member yet?
Click here to register.

Forgotten your password?
Request a new one here.
Shoutbox
You must login to post a message.

renatoa
03/29/2024 1:55 AM
Soubriquet and kaicee18, welcome to forum

renatoa
03/25/2024 12:38 PM
coffee drink, Ramper?

allenb
03/24/2024 12:10 PM
Siam3D Welcome

renatoa
03/23/2024 3:02 PM
welcome cup, andrewg

allenb
03/21/2024 5:10 PM
TRH629 Welcome

In Memory Of Ginny
Donations

Latest Donations
dmccallum - 10.00
JackH - 25.00
snwcmpr - 10.00
Anonymous - 2.00
Anonymous - 5.00
Users Online
Guests Online: 3

Members Online: 0

Total Members: 8,167
Newest Member: Soubriquet

View Thread

Who is here? 1 guest(s)
 Print Thread
Adventures in fluidbed roasting
allenb
So, after doing various experiments with controlling my small non-fluid, hot air roaster, via hot air temp, I want to share some results that I find very interesting and again is additional proof that conventional wisdom on what profiles are likely to give you best outcome are only suggestions and should never be hard fast rules for how you roast.

Example 1.
A Guat Huehue purchased in May this year has given me anywhere from ok to not so good outcomes in drum and fluidbed using various profiles that had always given me good results with many other coffees. I was ready to toss the greens into the trash until I did some additional roasts using the simple two stage approach I have been playing with for a few weeks. The stages were as before, 320F hot air inlet temp for 4 minutes and remainder of the roast at 510F. This resulted in the coffee ramping fairly quickly up to around 260F bean temp by minute 3 and basically staying at that temperature up to minute 4. The remainder of the roast at a hot air temp of 510 resulted in the bean temperature ramping very quickly to 400F where 1st crack started at minute 6:30 and ended roast at 7:35. Cupping results the next morning were excellent with nice balanced acidity, body and complexity. Probably the best quat I've cupped in years.

Example 2.
Uganda Bugisu purchased around the same time period was also getting close to being tossed due to not only having a "green" under developed taint but also with a pronounced woody note no matter how I roasted it using lots of variations of the classic curve. Using the aforementioned simple two stage approach, the coffee was remarkably better and very drinkable with much less earthy woody character and much improved complexity and body.

Another bit of info that proves that even slight variations can change a roast from ok to excellent.
In an attempt to see if what I accomplished above could be improved on, I made one small change by ramping the heat from 320F to 400F from minute 3 to minute 4 in order to prevent the coffee from stalling at 250F from minute 3 to minute 4. I was sure this would bring about an even better cup but was shocked to find it had degraded the cup and totally lost the body and complexity.

If someone had approached me earlier this year and told me they were getting great results with this method, I would have immediately written it off and would have been comfortable believing the person was unable to know what coffee is supposed to taste like and couldn't possibly be getting good results.

So, please never get comfortable with how you are now roasting and always be willing to try different roasting strategies. As many have stated, no one ever masters coffee roasting even though we get close from time to time. Of course, this is what keeps us drawn to the craft isn't it?
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
CK
Nice candid post Allen. I'm about five years into learning how to roast and have found that numbers, times, temperatures and curves do not always end in a great-tasting cup... especially on experimental machines. There are just too many variables involved with the process (from harvest to cup) to be stuck with one method of roasting. Like you, I've found that unconventional experimentation can bring great results too.
 
allenb
Hi CK. "Too many variables" yes indeed!

An outcome of my findings is I will now include taking any new green purchase through both the short hot air roast and a classic drum roast regimen to see if the bean favors one or the other and to also discover if I'm missing any flavor notes in either of them.

A Kenya Kigutha I picked up recently comes out really nice on both machines but displays very different sets of complex flavors using each method. Using the two roasters, it's almost like getting two different coffees from one purchase.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
CK
Roast Curves Screenshots; for a visual aid to see what happens with BT using this method of roasting

I've roasted two 600-gram batches, back to back, with a newly built hybrid roaster (drum/hot-air). First I used a generic stepped heater power profile, starting at 76% then stepped to 84% at 1 minute intervals. Second I tried the Schulman method you referred to in your post, with this machine at 50% power for 4 minutes then to 84% till the end of roast at 216C. It certainly looks off, but I'll cup them in a few days to see how greatly they differ, and just how good/bad they may taste.
CK attached the following images:
costa_rica_tarrazu_schulman_profile.jpg costa_rica_tarrazu_heater_power_profile.jpg

Edited by CK on 09/06/2021 7:23 PM
 
renatoa
Schulman advise about "ramp to...", not "step to"

Some 5-10-15% steps from 50 to 84 would made things look better, in the eye of a pixel
level graph peeper Grin
 
CharcoalRoaster
CK, do you have a build thread for your hybrid roaster?
 
CK

Quote

renatoa wrote:

Schulman advise about "ramp to...", not "step to"

Some 5-10-15% steps from 50 to 84 would made things look better, in the eye of a pixel
level graph peeper Grin


True, ideally stepping the heat would have looked better graphically and followed the recipe, but I didn't want a long roast, lagging 15-20 minutes.

How long should each percentage of ramp last from start to finish? I could program another batch to show a real-world graphic of what it would look like in that scenario... at least on this roaster.Grin
 
CK

Quote

CharcoalRoaster wrote:

CK, do you have a build thread for your hybrid roaster?


No, I didn't create a thread for this roaster.
 
renatoa

Quote

CK wrote:

Quote

renatoa wrote:

Schulman advise about "ramp to...", not "step to"

Some 5-10-15% steps from 50 to 84 would made things look better, in the eye of a pixel
level graph peeper Grin


True, ideally stepping the heat would have looked better graphically and followed the recipe, but I didn't want a long roast, lagging 15-20 minutes.

How long should each percentage of ramp last from start to finish? I could program another batch to show a real-world graphic of what it would look like in that scenario... at least on this roaster.Grin


Nope, not lagging anything, same levels, but at least an intermediate 70% step at 2:00, where RoC graph seems to start slowing its increase, should smooth a lot the drying phase of the graph, eliminating that RoC bump.
 
CharcoalRoaster

Quote

CK wrote:

Quote

CharcoalRoaster wrote:

CK, do you have a build thread for your hybrid roaster?


No, I didn't create a thread for this roaster.


Would you be willing to offer a short description and/or photos? I'm curious that's all... thanks!
 
allenb

Quote

renatoa wrote:

Schulman advise about "ramp to...", not "step to"

Some 5-10-15% steps from 50 to 84 would made things look better, in the eye of a pixel
level graph peeper Grin


So, it may be good for me to clarify what I meant by me saying I was using the suggestions by Jim Schulman.

As renatoa pointed out, Jims suggestion was to control ones roast via ET instead of BT since reading BT is unreliable and his preferred method was, as pointed out by renatoa, to step the temperature in ET or "ramp" it in order to more smoothly bring the beans up in temperature versus making abrupt, large increases.

The only reason I mentioned his method as relating to my experiments was in my using ET (hot air inlet temperature) for controlling my roast instead of BT. And, the only reason I've been sticking with a two step ramp instead of a multistep smoother approach is due to finding my coffee cupped better with the simple two step method.

If anyone wants to see if they get similar results as I have been getting, they will need to find ET temperature settings in two steps that allow the beans to be at somewhere between pale green and beginning of yellowing by end of minute 4 and to be at beginning of first crack by minute 7 or up to 7:30 and continue on for another 60 seconds hitting a light roast at completion. One should find a temperature for the second step that accomplishes this without having to adjust it between the beginning of 2nd step and end of roast. Depending on your hot air roaster's roasting dynamics (air flow, degree of agitation etc.) you may end up with widely differing hot air inlet temperatures to accomplish the profile in the time frame described.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
pisanoal
AllenB can you post up what one of these profiles looks like? I'm curious to try this myself.
 
CK
Curiously the profiles in post #4 cupped quite similarly. Always learning...

Here's another concept from Nestle. They roast the first part of the process, and then when you're ready you finish the roasting on your machine. See paragraph 2 under the video on this page...

https://www.nestl...ut-of-home
 
renatoa
This is indeed a mind blowing approach... Shock demolishes all we think we knew about the roasting process...
 
allenb

Quote

pisanoal wrote:

AllenB can you post up what one of these profiles looks like? I'm curious to try this myself.


I'm not charting my roasts but the next roast I do, I'll jot down the bean temps at 30 sec intervals and I'll report back.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
allenb

Quote

allenb wrote:

Quote

pisanoal wrote:

AllenB can you post up what one of these profiles looks like? I'm curious to try this myself.


I'm not charting my roasts but the next roast I do, I'll jot down the bean temps at 30 sec intervals and I'll report back.


Ok, here's the bean temperature time/temps for the two step hot air roast with hot air setpoint of 320F for 4 min and then 510f for the remainder. Realize that your roaster's thermodynamics might require a totally different pair of temperatures to get in the neighborhood of the profile. And, as you can see, it's an odd profile with absolutely no elegance. Also, one can really see the naturally occurring declining RoR when using fixed temps instead of ramping. I don't think there is any accuracy in my bean temps during first crack and up to end of roast but that is what the thermocouple is reporting. One can see the huge affect that the vapor release causes during 1C in trying to read the beans.

30 157F 69.44C
1:00 197F 91.66C
1:30 213F 100.55C
2:00 224F 106.66C
2:30 240F 115.55C
3:00 245F 118.33C
3:30 250F 121.11C
4:00 252F 122.22C
4:30 300F 148.88C
5:00 339F 170.55C
5:30 364F 184.44C
6:00 381F 193.88C
6:30 397F 202.77C
6:42 400F 204.44C first crack start
7:00 403F 206.11C
7:30 407F 208.33C
7:42 412F 211.11C end (first crack slowing down)
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
renatoa
A suggestive graphic representation of "huge affect" of the FC on hot air evolution.
Roast is done a hot air machine, not a FB, but the shape apply to FB also.

Beware, if temperatures range isn't a good hint: the data plotted is of air, not beans.
...
renatoa attached the following image:
0510_eth_yirgacheffe_gr_1_cherry_red.png
 
renatoa
Regarding ramp vs step debate, and the "how many steps" question, is hard to find an answer to this by trial end error instead performing a step response test of your process.
https://en.wikipe...p_response
Having such test done, it's easier to simulate a ramp composed from several small steps, as in the attached image.
...
renatoa attached the following image:
3dfeb315ff28c7e1c4117f6bf283836bede474dd.jpg
 
nickr
Allenb,

I know this is an old thread but are you still using this method? Did your results continue to be good? I like the simple approach since mine roaster is manual.
 
allenb
Hi nickr. I'm still using the method but my Hearthware Gourmet hot air roaster died (blower wheel committed suicide) and was not the kind of fan that can be rebuilt. I've since made some mods to my old Freshroast and am using close to the same two step profile although the second phase temp is way different due to the Freshroast being able to pull off the Maillard and develpment phase at 480 F instead of 510 F and still hit end of a light roast at around 7 minutes. The roasts aren't quite as good as the Hearthware Gourmet was giving me but is close.
Since changing to the Freshroast roaster, I did a lot more experimentation with comparing complex multi staged ramps versus the 2 step and never found an improved cup with the multi stage.

Something I did find while roasting lots of different coffees since starting the thread is that there are green coffees out there that do not like a short 7 minute roast and exhibit woody/earthy or other under developed taints that are much less apparent when using a 10 minute or longer profile. Fortunately, it's only been maybe 10% of my green purchases so far.

An important thing to nail down when setting your 2nd stage temperature is that you end up with a high enough air temp so that the roast doesn't ever drop lower than around 8 F per minute from first crack onward or you will bake it and lose all the high notes. With the temp I'm now using, for a very light roast, the coffee never takes more than 1 1/2 minutes from start of first crack through end of roast.
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
nickr
thanks for the fantastic reply!

couple of follow up questions:

I roast mostly with my home-built roaster but also have an SR800. What settings do you use for your 2 stages on the Freshroast?

Is there a specific origin that doesn't agree with this method or is it random? I mostly roast Indonesian coffees. I enjoy the results using this method, but I'm new and may not know how to recognize any defects.

If you wanted to reduce acidity, which stage would you modify, 1st or 2nd?
 
allenb
Edit Shock

I posted 2 minutes of soak at 360 in the first sentence below but this is not correct. Should read 3 minutes soak so I'm changing it as I write this.

I pulled all controls out of my freshroast and control the heater via a Fuji PXR4 ramp soak controller with zero ramp time to first stage temperature of 360F and 3 minutes soak. Next stage is 480 with zero ramp time and plenty of extra soak time to ensure it doesn't cycle down during the roast.
I threaded a small thermocouple in through the side of the heater section and up through the mica board to get a good accurate ET read which is connected to my Fuji. I did an autotune for PID setting and works reasonably well with little over/undershoot.
The fan is controlled via a 3 amp staco variac. I don't believe it's possible to correctly control a freshroast without scrapping the factory controls. I have found it necessary to make continuous micro adjustments to the airflow in order to keep optimum lift throughout the roast which also allows the freshroast to efficiently ramp to the next setpoint with as little lag in reaching the next setpoint as possible. Due to the freshroast having barely adequate kw, at least in my early model, when the second stage setpoint starts (480F), it takes roughly 2 minutes before ET actually reaches 480F from it's previous setpoint of 360F

As far as coffees that didn't like the 7-8 min roasts? It is random. I think it is a result of errors in processing the cherries where there is fermenting problems or other processing errors somewhere in the chain. This happened with a Kenya recently and also with a Guat. Haven't had it happen with any of my Ethiopians or any other Kenyas.

One thing of note that I didn't expect to find in my tests. I now roast all new coffees received through the freshroast and 1lb gas drum roaster for comparison and without exception, all coffees exhibit much more of the high notes (floral, spice and other) when roasted in the freshroast fluidbed if kept to an 8 min or shorter roast. I've tried doing an equally short profile in my drum roaster and was not a good roast although, with just the right bean, I have had on occasion a very tasty coffee from the drum roaster using an 8 min profile but never produced the same level of high notes.

On reducing acidity, the only method I've ever used is to increase roast level in very small increments until you have just the right amount. Where most people go wrong is they add too much time on the next roast and end up killing all of the acidity and end up with a lifeless cup.
Edited by allenb on 10/18/2023 4:53 PM
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
HarryDog
Hey allenb, do you have a pic of the probe placement?
By the sounds of it right above the heating element?
My probe is a bit high and I have some trouble with 250F charge temps.

Right now I have been using 200F charge temps on natural and Honey Process beans.
Also slower roast times.
 
allenb

Quote

HarryDog wrote:

Hey allenb, do you have a pic of the probe placement?
By the sounds of it right above the heating element?
My probe is a bit high and I have some trouble with 250F charge temps.

Right now I have been using 200F charge temps on natural and Honey Process beans.
Also slower roast times.


Not sure how your freshroast is configured but with mine, the heat exits from the perimeter surrounding the bottom of the metal portion of the roast chamber when it's in place. The bottom of the discharge area is a flat micaboard plate which I drilled a micro small hole allowing the cloth insulated thermocouple wire to come up through for about an inch sticking relatively straight up. With this placement, the air surrounding it passes up and across the tc on it's trip upward.
The image showing the whole roaster will require some explanation. I pulled the top plate with fan/heater assembly out of the freshroast body and got rid of the body. The top plate fit perfectly into my out of commission hearthware gourmet white plastic base and I bolted a work box to the side of it to house my Fuji controller and staco variac.

A start temp of 200F is too low from my experience and doesn't give the coffee the initial kick of heat that seems to enhance most roasts.
allenb attached the following images:
img_0623.jpg img_0622_2.jpg

1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
HarryDog
Thanks Allenb, the base I use is out of a popper and when I used a probe from the top I always thought the temps were right. I closed it off to roast inside so I put a probe on the side and think the temps are just wrong at pre-heat temps to use it as an ET probe.

Was thinking of putting a probe at the base on the side so I don't put it right into the middle of my heater which I don't think it would last?

If I don't drop the temp for Naturals and Honey processed beans I normally burn the chaff. I would have to test it again but I suspect the temp at 200F is closer to 250F or more?
 
Jump to Forum:

Similar Threads

Thread Forum Replies Last Post
New to roasting. Could you help me with roasting advice? Roasting Coffee 24 02/18/2024 10:45 PM
Hottop KN-8828P-2K - Roasting Profiles HotTop Roaster 7 02/09/2024 7:05 AM
New to roasting and could use helpful advice w/SR800 roasting. Roasting Profiles 8 01/29/2024 9:36 AM
Re roasting underdeveloped coffee beans Roasting Coffee 5 01/02/2024 11:17 PM
Measuring Bean Temp in a Fluidbed Fluidbed Roaster 11 11/09/2023 3:33 AM
Homeroasters Association Logo, and all Content, Images, and Icons © 2005-2016 Homeroasters Association - Logos are the property of their respective owners.
Powered by PHP-Fusion Copyright © 2024 PHP-Fusion Inc
Released as free software without warranties under GNU Affero GPL v3
Designed with by NetriX
Hosted by skpacman