Buyers Remorse - Kaleido M10 Astringent Roasts - Page 3

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
Agtron70 (original poster)
Posts: 15
Joined: 2 months ago

#21: Post by Agtron70 (original poster) »

PaulTheRoaster wrote:Do you need to use PID control? Most drum roasters use power instead (and wouldn't oscillate from 100% to 30% and back). That could possibly be the issue.
Automated roasting is a very important feature for me. If I were to eliminate that as a purchase criteria, I probably wouldn't have purchased a Kaleido M10.

I don't really care for the idea of manual roasting and altering controls during the roast but that is where the next phase of testing is headed.

User avatar
yakster
Supporter ♡
Posts: 7334
Joined: 15 years ago

#22: Post by yakster »

The automated roasting will be helpful once you develop a good working roast profile, probably manually.
-Chris

LMWDP # 272

Agtron70 (original poster)
Posts: 15
Joined: 2 months ago

#23: Post by Agtron70 (original poster) »

drgary wrote: Without time at the moment to consider driving the roaster with Artisan and a PID (have you thought of reaching out to Kaleido and showing them your curves? They may suggest a change to PID settings)......

.....By comparison, your roast looks pretty rough, especially with the dip in bean temperature rate of rise early on. Uneven ROR like that will tend to develop the coffee unevenly.......
I have reached out to Kaleido in China via email on another matter - radio silence. I need to shake that tree again and try to get a response.

Re: The roughness of the graph, I have not turn on any smoothing post roast, so the rough graph is a depiction on basically raw data.

I would very much like to improve the smoothness of the raw data, moderate the power activation during PID control, and finally have classic decreasing RoR graph.

More science project awaits.

Agtron70 (original poster)
Posts: 15
Joined: 2 months ago

#24: Post by Agtron70 (original poster) »

yakster wrote:The automated roasting will be helpful once you develop a good working roast profile, probably manually.
I hope you are correct. I shall find out.

I am concerned that a good manual roast profile cannot be PID controlled to the same level of success. I believe the PID control on the Kaleido M10/Artisan is entirely software based by Artisan, not a true robust industrial type hardware based PID - I could be wrong on that.

I am also concerned that the nature of the infrared heat design - that is the way the heating is laid on the environment and the beans - has certain unpleasant and unaddressable artifacts.

PaulTheRoaster
Posts: 261
Joined: 18 years ago

#25: Post by PaulTheRoaster »

Agtron70 wrote:Automated roasting is a very important feature for me. If I were to eliminate that as a purchase criteria, I probably wouldn't have purchased a Kaleido M10.

I don't really care for the idea of manual roasting and altering controls during the roast but that is where the next phase of testing is headed.
I agree with you-I have a Bullet and use recipes exclusively, but they are simple power-based recipes. PID drum roasting just seems like a bad idea to me.

Agtron70 (original poster)
Posts: 15
Joined: 2 months ago

#26: Post by Agtron70 (original poster) »

drgary wrote:A fluid bed roaster like a Kaffelogic or my IKAWA Home drives a lot more air through the beans than on my drum roaster.
For a given bean that you roast on both your drum roaster and your Ikawa roaster, are the FC temps the same on both roasters?

If not, how do they typically differ?

Rytopa
Posts: 228
Joined: 7 years ago

#27: Post by Rytopa »

I finally see where the issue is at, PID roasting on the M10 is a big no no, the wide swings in % power is causing a lot of problems for the beans, the machine behaves more like a traditional drum roaster vs Kaffelogic which is alot more nimble and sensitive. PID is good for setting the Charge Temp and that is about it.

Start doing traditional manual roast adjustments and understand the characteristics of this roaster, automation can happen but only after you have gotten a good roast profile manually, Artisan can replicate that roast profile for you using play back events by BT



This setting makes adjustments to the power setting based on the BT Temp, example, once it hits a certain temp , the power adjustment is made, doing this way forces the machine to only have a linear decreasing power % adjustment without the wild swings of the PID.
★ Helpful

Agtron70 (original poster)
Posts: 15
Joined: 2 months ago

#28: Post by Agtron70 (original poster) »

Rytopa wrote:...PID roasting on the M10 is a big no no, the wide swings in % power is causing a lot of problems for the beans....
My rough understanding of how the PID is supposed to work is that the PID uses the background BT as the SV and tries to manage the roasting BT as the Process Variable (PV), and through some nice math it supposed to manage the power modulation to produce a roast that is consistent with the background. Not sure why the power is fluctuating in the large way it is. Even though the BT curve matches the background pretty well it appears that the roasted product is marginalized by the histrionic RoR.

This PID roasting journey went off the rails. I would have probably have picked an Aillio Bullet had I knew how the trip was going to end up.

So now I am going to try to do exactly what you and others have suggested by starting with a manually derived curve and then possibly automating with Artisan alarms or other means. Making lemonade out of lemons.

Rytopa
Posts: 228
Joined: 7 years ago

#29: Post by Rytopa »

M10 is predominately driven by Infrared, which will require drastic adjustments to the power as the beans get darkers, as it absorbs more heat from the heat source, given that thermocouples have an inherent lag , coupled with the above phenomena, corrective heat adjusts have to be made before its shows up in the graph, which the PID will not be able to capture.

The Bullet which i have owned before also suffers from this greatly, being primarily a conduction heat source, it is nowhere capable of using the nanuces of a PID driven system to follow curves.

For such systems the best is still an temp based event adjustment, say we know that once 180c (critcal temp where by beans have lost most of the moisture and is very sensitive to heat) is reached, we can tell the machine to do bigger 10% adjustment.

How i usually set up is 80% charge power, airflow max thruout, 160c -2%, 165c -2%, 170 -2% 175c -2% 180c -5% 185 -5% -190c -5%, at this point as you are nearing FC, you will likely back off from the heat adjustments and let it run thru FC, 45 secs after FC, -10% followed by -10% every 5c

This profile should generally give you quite good results which you can find tune , after which you can automate it.

Agtron70 (original poster)
Posts: 15
Joined: 2 months ago

#30: Post by Agtron70 (original poster) »

Rytopa,

Thanks for the roast power adjustment advice.