Learning to roast on Quest M3s

#1: Post by **Cwilli62** » October 16th, 2020, 9:59 pm

Hello All,

I started my roasting journey on a Behmor 1600 and several months ago moved to a Quest M3s. I have an okay grasp on roasting practices, though I admittedly should do some more reading in books and on the web.

Anyway, I know there are basic guidelines for roasting, no matter the machine but I figured it would be nice to have a running thread on specifics to roasting on the Quest (along with general roasting info. if needed/desired to further overall understanding).

I've read through the following threads and find them only moderately helpful:
Help me get better roasts with my Quest M3 (with graphs)
Quest M3 guide and experiments
Trials, tribulations, and learnings roasting on a Quest M3
If anyone knows of other threads that may help, please let me know! These are the only ones that I've found.

The thought process of trying to learn how to better roast, specifically with the Quest, was brought on by:
1) My roasts not turning out nearly as well as I want them to/know they can
2) Recent discussion in this thread about M3s thermocouples: Quest M3S Thermocouples
The most recent posts in that thread about the usefulness of certain probes in the Quest, what they tell you about how this machine is reacting, and when/how you should utilize them is helpful information for me.

In essence I thought this might be a good jumping off point for myself as well as others. If you're an experienced Quest M3 coffee roaster, a newbie Quest user that is likewise inquisitive, or an experienced roaster in general with knowledge to offer, feel free to share. I'm open to looking at/posting graphs of roasts, specifics of bean/batch size/method, etc. Just wanted to throw this idea out there first and see if there is enough interest to have a dialogue on the topic.

#2: Post by **Peppersass** » October 17th, 2020, 2:43 pm

This quote is imported from the Quest M3s Theromcouples thread:

appfrent wrote:... happy to say that controlling electric heating is a non issue (I only roast 200g batches, so that's the caveat).

Two questions:

1) What's your general sequence for controlling heat and air?

I usually start with MET at whatever maintains the target charge temp, and fan at just enough to move some air (25% to 30% of max). I increase MET at TP to, say, 500F, and boost fan speed to about 40% at peak ROR. I boost fan speed to 50% about a minute or so before DE. About 90 seconds before 1C I lower MET to about where it was at charge, usually in one step but sometimes in several steps. I hit max fan just before drop to clear any remaining chaff.

2) Have you experimented with larger or smaller batches? If so, what were your results?

Until recently, my Quest had the original M3 drum with a single row of perforations drilled around the back circumference. I started out with 175g batches, mostly because I was learning to roast and didn't want to waste more coffee in case a batch turned out badly. That worked reasonably well, but later I painted the drum black (hoping for more heat retention) and reduced the batch size to 150g so I could lower MET and compensate for the limited airflow. Generally speaking, I don't think those batches turned out as well as the 175g batches.

Now I have an as-yet untried M3S drum, and I'm thinking that the improved airflow will likely require pushing the batch size back up in order to keep MET down and avoid the roast moving too fast. I've read some posts that suggest a larger batch size is a benefit of upgrading the drum. Some seem to be happy with 225g batches (full capacity.)

#3: Post by **appfrent** » October 19th, 2020, 12:56 pm

I would be happy to contribute to discussion. I am relatively new (<year) to roasting (regret) so I have a long way to go. However, my results are already same or better than what I way paying $20-$30/pound for.

Peppersass wrote: 1) What's your general sequence for controlling heat and air?

My basic sequence is as follows.
Fan: At 1 throughout the roast. 9 just before first crack.
1) Preheat the roaster well. I set my heat such that it hangs around BT of 450F for 5 min or so. When I am ready, I cut the heat to 0 and let BT drop to 390F. Open the bean loading chute. Then, I blast the heat to Max and let the ET ramp back to 420F (400F for medium and dark roasts). If my heat injection is what I want MET is always at 550F at this point.
2. Load the beans.
3. Close the bin loading chute at or before 300F.
4. First heat cut between 300F to 330F depending on bean and roast type.
5. Second heat cut between 360-390 depending on bean and roast type.
6. Blast with max heat injection of 20s 5-10F before the first crack.
7. Cut the heat to 0 once first crack is underway.
8. Once the first crack has started, start a 2 minute timer and stop looking at numbers. Focus on beans. Smell and Watch. Still learning this part.
9. If confused/undecided, drop at last time that worked best or at most 2 minutes (or whatever upper limit you want to set) after the start of first crack (for light roast).

For darker roasts, I start with lower heat and end up with more heat momentum at first crack to carry through (shallow and longer).

General notes:
1) The absolute temperatures are meaningless. Every setup will have different numbers for sure.
2) Learn to anticipate the time lag between temperature change/adjustment vs MET/BT reading.
3) On my QuestM3S heat cuts needed to be far more drastic than I anticipated.
4) Don't read/interpret the time and temperatures alone, always together.
5) There is no maxima or single value to optimize. There is a band of acceptable values. This band will provide you a range of good flavors. Different, but all equally good. The band could be narrower or wider depending on the coffee bean. Depending on who you are, this can be the best thing in the world or an utter nightmare.
4) Simplify workflow and variables as much as possible. I was doing zillion heat cuts; in the end I learned that just 2 are sufficient. I was too worked up about the following certain curve. Also turned out to be unnecessary. At least for me, a simpler workflow allows me to focus my attention at a big picture and be ready for the moment that matters most, the decision to drop.

Peppersass wrote: 2) Have you experimented with larger or smaller batches? If so, what were your results?

No. I picked 200g because that was what was commonly used here. Fortunately, it works best for my situation. I do 4 X 200g every alternate weekend. They last me 2 weeks. I time roasts such that I still have half a week worth of previous beans left. I always roast 3 different light roasts and one medium roast. I really enjoy the variation. Sometimes, I do the same bean with tiny differences and its great to have two different flavor profiles from the same set of beans.

#4: Post by **yamhill** » October 19th, 2020, 2:53 pm

appfrent wrote:
My basic sequence is as follows.
Fan: At 1 throughout the roast. 9 just before first crack.
1) Preheat the roaster well. I set my heat such that it hangs around BT of 450F for 5 min or so. When I am ready, I cut the heat to 0 and let BT drop to 390F. Open the bean loading chute. Then, I blast the heat to Max and let the ET ramp back to 420F (400F for medium and dark roasts). If my heat injection is what I want MET is always at 550F at this point.
2. Load the beans.
3. Close the bin loading chute at or before 300F.
4. First heat cut between 300F to 330F depending on bean and roast type.
5. Second heat cut between 360-390 depending on bean and roast type.
6. Blast with max heat injection of 20s 5-10F before the first crack.
7. Cut the heat to 0 once first crack is underway.
8. Once the first crack has started, start a 2 minute timer and stop looking at numbers. Focus on beans. Smell and Watch. Still learning this part.
9. If confused/undecided, drop at last time that worked best or at most 2 minutes (or whatever upper limit you want to set) after the start of first crack (for light roast).

This is quite similar to my approach with a couple of differences, but before I note the differences, it may be important to recognize that the Quest M3 has some different drums -- and measured temperatures are likely different from machine to another based on nitty thermocouple and machine differences. One of the primary insights implicit in the process above is the use of the doors. The doors can be used as an absolute binary switch on the fan. Also, my machine has a solid drum and so air only seeps around the end. When my machine is idling ET and BT measures are very similar as they are only a couple of inches apart and the air flow is fairly low.

Now some differences.
I typically charge at 400 ET (BT reads pretty much the same).
I charge into a roaster that is running at 1000-1200 watts -- measured with a kill-a-watt.
I leave either the bean-loading flap or back cooling tray flap open until BT gets to around 300.
I increase the fan throughout the roast.
I sometimes cut the heat at first crack, but usually I just increase the fan.

Here's my general process for city to city-plus roasts.

Here are a few other notes: I'm *finally* fairly comfortable making adjustments on the fly. I have 1.5mm probes and get fairly quick feedback. I find the fast probes helpful as I roast a wide variety of coffee, and I roast in my garage where the roasting environment can vary from maybe 50-85f.

My experience with the Quest M3 is that it takes 30 seconds or more to "see" an increase in heat. This is important to note especially if the roaster is coasting without any heat in the elements. If the fan is running and the elements are cool, the roaster will continue to loose heat and not see any effect from the increased heat setting for about 30 seconds -- at which time it will start to gain back any heat pulled out of the roaster during that time. -- And, the elements are large and hold head for a while. If you are wanting to cool things down quickly and you cut the heat and increase the fan -- the roaster will respond by blasting the beans with the heat from the heat stored in the elements and other roaster parts for 30 seconds or so. If you're really desperate to reduce heat, you can crank the fan and pull the trier.

I have so much to learn..., but I'm mostly enjoying the process.

John

#5: Post by **Peppersass** » October 19th, 2020, 3:44 pm

Great stuff. Some comments:

1. I don't know how much air a fan setting of 1 moves because my Quest doesn't have a fan pot. It's setup to control the fan via an Arduino/TC4 interface, so I select fan speed percentage in Artisan. For me, 25% barely moves the fan, so I use 30% as the lowest practical speed. With the old M3 drum I found that I had to increase fan speed to 40% at peak RoR to prevent RoR from dropping steeply, then up to 50% a minute before DE to keep ROR from going to zero during 1C. In other words, I needed more air to circulate heat. I'm looking forward to seeing if I can use lower fan speeds with my new M3s drum.

2, Interesting that you leave the bean chute open for quite a while -- even to DE. That must improve the air circulation, which perhaps explains why you can get away with such a low fan speed. Perhaps closing the chute serves to lower heat indirectly even before you reduce heat directly.

3. I think some people leave the chaff door open early in the roast for the same reason.

4. I try to keep MET at 500F or lower, so probably more air is needed to do that with the old M3 drum.

5. Normally, I don't cut heat between 300F and 330F, but it might be a good thing to try as I've heard 90 seconds between DE and Brown is ideal, and my roasts usually fall short of that.

6, The blast of max heat for 20s 5-10F before 1C is new to me. Is that to avoid an RoR crash? Usually my RoR rises just before 1C, so I'm wondering if goosing the heat like that will cause a spike.

7. Zero at 1C is new to me as well, though that might go well with the blast of max heat.

8. As for focusing on Smell and Watch, with the M3 drum I can't use the tryer because removing it causes substantial change in airflow. I have to sniff around the exhaust, which is connected to a vent, and look through the sight glass. I'm hoping the greater airflow with the M3s drum will make it possible to use the tryer.

9. I agree that simpler is better. I'm hoping with the new drum I can set one constant airflow for the entire roast and do minimal heat changes.

#6: Post by **Dokkie** » October 19th, 2020, 4:32 pm

I'd actually recommend maxing out the airflow for the entire roast. The airflow is pretty low even at the max. I've charged anywhere from 150 to 400 grams (extreme, drum will barely hold it). Best results with a 150 to 250 gram charge. Funny enough even the 400 gram charge I can get to 1C in 8 minutes, albeit with a bit higher charge temperature. The elements are pretty strong.

I've attached 2 graphs, Ywagwan is 250 grams and 215 roaster scope is 150 grams.

#7: Post by **yamhill** » October 19th, 2020, 11:45 pm

Dokkie wrote:I'd actually recommend maxing out the airflow for the entire roast. The airflow is pretty low even at the max. I've charged anywhere from 150 to 400 grams (extreme, drum will barely hold it). Best results with a 150 to 250 gram charge. Funny enough even the 400 gram charge I can get to 1C in 8 minutes, albeit with a bit higher charge temperature. The elements are pretty strong.

Thanks for suggesting this. I'll give it a try -- or, at least I'll think about it between now and my next roasting session. I wish I understood physics better. It seems like this would/could apply the same energy, but transfer more heat via convection and less through conduction -- than lower fan speeds. I do have a some questions. What's the ambient temperature while you're roasting? Is your roaster modified at all? What type of drum do you have -- solid or perforated?

#8: Post by **Dokkie** » October 20th, 2020, 1:34 am

I use the M3S, no modifications (so perforated), my ambient temperature is around 20c most of the time.

October 20th, 2020, 9:53 pm

There's a lot of good advice and information here already! I'm glad that others are willing to share their experiences. I've been with M3s for about 6 months now. I have tried a variety of things, but 2 months ago I realized that I need to keep it simple and have a baseline process so that I can actually learn what tweaks produce which results. My roasts were okay at first, then straight trash because I was trying to get cheeky with "Nordic style" roasts and doing a poor job at it, and now they are back to okay. Nothing amazing yet for me.

Definitely a LOT to learn still. It is fun and extremely frustrating at the same time. A good hobby to have.

I pay attention to BT mainly and have recently been paying more attention to MET. I find it hard to make MET do what I want it to do without BT going astray. I never pay attention to ET probe, because as others have mentioned, mine follows pretty closely with BT probe.

Anyway, now my process seems similar to Dokkie's. The following general processes are what I have been playing at for the last couple of months...

Fan: I was keeping my fan speed at 5.5 the whole roast (which is a little bit above the lowest setting), but I may take the suggestion to try maxing it out for the whole roast.

Batch size: Typically I roast two, 114 g batches every week or two. I was doing the smaller batches so that I could get more roasts in but it seems like this may be too small according to what y'all are roasting. I can up it to two, 150-175 batches maybe? Or just do one 230 g roast and call it a day.

Temps./power: I charge around 280 F with these smaller batches. The problem is that since they are so small, everything happens so quickly. Perhaps another reason to increase batch size. Heat reductions starting around 4 min. and at 1-1.5 min intervals, depending on how things are looking. Quicker heat reductions toward the end of the roast. Recently I was experimenting with "soaking" right after charge for 45 sec. Not sure if I like that or not. I have a Kill-A-Watt but haven't used it yet. I suppose it would help to keep things more standard and be better able to converse with others about heat settings if I used it. Right now I have been starting on 8.5 A setting at the beginning of the roast and getting down to 6 or so by the end of the roast. Typical drop temp for me is 400-410 F, about 45 sec-1.5 min after 1C, depending.

Other tweaking and fiddling: I found that opening the bean chute or the chaff collector add another variable that I don't want to mess with. Likewise, I found that removing the trier throws the probe readings a bit, so I usually don't mess with it.

Here are a couple of graphs of recent roasts. I still don't have much idea on how to change the roast depending on the specific bean.

#10: Post by **appfrent** » October 21st, 2020, 2:01 am

Judging from the responses here, fan settings on different roaster models/units may be very different. For me, there is just a modest increase from 1 to 9 and 1 already has a very robust airflow. Hence the reason to leave it alone and increase only at the end. Everything else fits the general pattern with minor nuances.