Roasting ain't easy - not even for Marshall Hance - Page 5

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
CarefreeBuzzBuzz
Supporter

#41: Post by CarefreeBuzzBuzz »

edpiep wrote:Gonna need to get more practice greens :roll:

Check out Genuine Origin. 65 lbs - $10 shipping and often free. Beans in the high 80's. We have had quite a few we like from their and great for practicing on.
CarefreeBuzzBuzz
___
Artisan Quick Start Guide
http://bit.ly/ArtisanQuickStart

User avatar
Peppersass
Supporter ❤

#42: Post by Peppersass »

Seems to me that roaster type, probe type and probe position greatly influence our perceptions of what information is useful.

My Quest M3 has four probes:

1. A 1/16" BT probe in the door next to the sight glass, which puts it low in the drum on the side where the beans congregate.
2. An 1/8" Alt-ET probe on the opposite side of the drum, about halfway up (labeled Alt-ET.)
3. A 1/16" MET probe between the drum and outer shell of the roaster (labeled MET but treated by Artisan as ET.)
5. A 1/16" Exhaust probe in the exhaust tube.

Here's a graph of a roast that shows the information provided by each probe:

Image

Several important points:

A. The heat in my roaster is PID-controlled by an Arduino based on the MET probe. I set a temp, and the Arduino throttles the heating elements to get there and stay there.

B. As you can see by how MET and Alt-ET react to my heat settings (the black PID line), it would be impossible for this to work with the Alt-ET probe. For example, you can see that when I increased heat after TP the MET gradually increased to the target setting while Alt-ET actually decreased, perhaps due to the cooling of the probe by the beans absorbing heat from the environment.

C. If the PID ran off Alt-ET, the heaters would have to be at 100% for a long time to move the Alt-ET line, which would surely have scorched the beans and/or screwed up the rest of the roast.

D. Unfortunately, I can't get a delta Alt-ET curve because to get Artisan to PID off MET I have to designate the MET probe as ET (though I can label it MET.) So the curve would show the MET RoR, which isn't useful. Eyeballing the slope of the Alt-ET curve suggests that its RoR doesn't correlate well with MET or BT.

E. Alt-ET and BT converge, as they do in most roast curves posted here, but unlike those posts my Alt-ET and BT are still pretty far apart near the end of the roast. The convergence I see for other roasters would seem to indicate that the bean temperature is getting close to the environmental temperature, which if used as a target for drop might be a useful indicator, but in my case the curves are still pretty far apart at that point and it's not clear to me whether some of the convergence is being caused by heat coming from the beans shown by the dramatic flick. On my roaster, I think the differences in probe size and position account for the gap between the BT and Alt-ET curves -- i.e., I suspect that two identical probes halfway up the drum on opposite sides would be much closer near the end of the roast, as seen in some of the curves posted here. But that still wouldn't work for Artisan controlling heat via PID.

F. The Exhaust curve doesn't tell me much at all about what's going on in the roaster. Could be the location or maybe because the M3's airflow isn't great.

G. Despite the flick, this roast tasted pretty good. It's a natural with a lot of blueberry flavor that I was trying to bring out. I got some of that, but there was too much astringency. Not sure if that was due to lack of development, the flick or both.

The bottom line is that what happens in my roasts correlates closely with temperature settings, which are clearly reported by the MET probe. If MET is too high at charge, I get tipping. If MET isn't increased after charge or a soak, RoR will decline too fast. If MET isn't dropped 1-2 minutes before 1C, BT will flick (as you can see in this roast, I needed to drop the heat earlier and probably more aggressively.) I'm not seeing useful information from the Alt-ET and Exhaust probes.

edpiep

#43: Post by edpiep »

CarefreeBuzzBuzz wrote: Check out Genuine Origin. 65 lbs - $10 shipping and often free. Beans in the high 80's. We have had quite a few we like from their and great for practicing on.
Thanks Michael, I have had some good experiences with GenOr. I will def check them out.

User avatar
MaKoMo

#44: Post by MaKoMo »

Peppersass wrote: B. As you can see by how MET and Alt-ET react to my heat settings (the black PID line), it would be impossible for this to work with the Alt-ET probe.
That black line looks to me like the SV (target) for the MET temperature not the actually heater power.
Peppersass wrote: D. Unfortunately, I can't get a delta Alt-ET curve because to get Artisan to PID off MET I have to designate the MET probe as ET (though I can label it MET.) So the curve would show the MET RoR, which isn't useful.
Why not use a symbolic formula to calculate and draw the delta Alt-ET from the Alt-ET data? It got quite simple to draw the RoR of any curve with the current Artisan v2.1

User avatar
Peppersass
Supporter ❤

#45: Post by Peppersass »

MaKoMo wrote:That black line looks to me like the SV (target) for the MET temperature not the actually heater power.
Yes, it's the SV setting, or an indirect version of it, not the heater power or duty cycle. The previous owner of my roaster setup the configuration and chose "PID" for the name of the curve. I can display the duty cycle but it doesn't provide information that I've found useful.

I say the SV is indirect because the "PID" curve is assigned in Devices to Arduino TC4 78. It's not clear to me whether Artisan is reporting the target temperature it's sending to the Arduino PID code on that channel or if the Arduino PID code is reporting the target setting back to Artisan on that channel. I would think it's the latter, given that the other configured channels are inputs (Alt-ET, Exhaust, etc.)

As I said, this was configured by the previous owner of my roaster. It functions well, so I haven't taken the time to dig deeply in to the Arduino configuration and firmware to figure out how it all works. I haven't even updated the firmware since I got the roaster a couple of years ago. First I have to make sure I have a copy of the currently loaded version in case something doesn't work right after the update. I do a fair amount of Arduino (and other) coding/troubleshooting, and enjoy it, but at this point I feel my time is better spent improving my basic roasting skills with what I have.
MaKoMo wrote:Why not use a symbolic formula to calculate and draw the delta Alt-ET from the Alt-ET data? It got quite simple to draw the RoR of any curve with the current Artisan v2.1
Didn't know I could do that. I'll look into it. There's a ton to learn about Artisan and I haven't had much time to delve into the documentation. Also, my math skills leave a lot to be desired, as do my roasting skills!

edpiep

#46: Post by edpiep »

I went ahead and made mods to my configurations before I pumped out two roasts showing a new perspective on my Artisan display. Overall it's simpler and less fussy. I did a test roast with some really old washed Mexican to get a feel of a "hotter start" roast then also did a roast of a nice Nicaraguan natural I have more of. My old "ET" line is now gone, my old "MET" line in now my new "ET" line and acts as one that many other roasters can understand. I realized my old ET line it pretty meaningless unless other wise proven useful. For now I am good with this new, more refined display.

I also really didn't care to "groom" my deltaBT at all on these. My goal was to try and reach MET and coast on that for as long as possible w/o skyrocketing out of control until the end of roast. The Mexican profile is just for feel, I can't objectively cup this coffee to make any good notes on this profile's characteristics b/c it just tastes like a wet, brown paper bag at this point. Roasts like a champ though :lol:

The natural is also just for feel, I went hotter and faster than an older profile I liked for this coffee. My deltaBT was way prettier on that one but had a minute extra roast time. Also note that the soak for the natural is not shown but was 1 kPa, 30 AIR @ charge for 1 min. I wanna see what this one tastes like after a few days, it fragrance smells like purple grapes and brandy though.

Mexico:
Image
Natural:
Image

User avatar
CarefreeBuzzBuzz
Supporter

#47: Post by CarefreeBuzzBuzz »

So why not mark MET (sorry laughing at myself here), cause you do a good job of keeping the new ET flat, and that way you will know the MET (sorry must have been stuck in this house for a long time.)
CarefreeBuzzBuzz
___
Artisan Quick Start Guide
http://bit.ly/ArtisanQuickStart

edpiep

#48: Post by edpiep » replying to CarefreeBuzzBuzz »

I would if I could. My old ET curve is hidden and that is what would be marked. Can I change what curve MET is marked on?

User avatar
CarefreeBuzzBuzz
Supporter

#49: Post by CarefreeBuzzBuzz »

Don't think so, but maybe Marko will tell us. Why can't you make the MET probe into the ET probe though.
CarefreeBuzzBuzz
___
Artisan Quick Start Guide
http://bit.ly/ArtisanQuickStart

edpiep

#50: Post by edpiep » replying to CarefreeBuzzBuzz »

Yes, that is an option. I haven't really investigated the length of my probe cables to ensure there would be enough length to switch probe locations but I think it would work.

It will probably be a project for this weekend. Gonna take a break from roasting for a day lol. Not like I have literally all day to do the project, which I do :shock: