Quest M3 bean mass depth

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
Peppersass
Posts: 3690
Joined: 15 years ago

#1: Post by Peppersass »

This is not so much about roasting as what goes on inside a drum roaster and how that affects BT probe measurements (in other words, I doubt that getting answers to my questions will make my roasts any better!)

It's my impression that the drum vanes in my Quest M3 are angled in order to route the beans toward the front of the drum so they'll come out of the drum and into the cooling tray when the door is opened. I'm also guessing it ensures enough beans will make it into the tryer trough when it's rotated upward, keeps enough beans against the site glass for the operator to see the color changes, and helps to ensure the BT probe is in the bean mass. All true?

Clearly, the position of the bulk of the bean mass for any given batch size is important for placement of the BT probe. I've seen lots of posts that talk about the bean mass position relative to the height and circumference of the drum, but not much about the depth of the bean mass into the drum and how that affects the BT probe.

I ask because my 2014 Quest M3, which rotates counter-clockwise, has two BT probes: a 2" x 1/16" probe in Alt BT position (immediately to the right of the sight glass) and a 4" x 1/8" probe in the OEM BT position (directly under the tryer.) The Alt BT probe extends into the drum about 1" from the back of the faceplate. The BT probe extends into the drum about 2 1/2" from the faceplate, as far as it will go without hitting the drum cross-members.

As you might expect, the two probes don't read exactly the same. The BT probe is usually about 80F higher than the Alt BT probe at charge, and takes about 1:30 to get to the TP, while the Alt BT probe gets to TP in 30 seconds (yes, I know the beans are actually rising, not falling to a TP.) When the BT probe hits the TP, the temperatures of the two probes is almost identical, and stays that way until about 30 seconds before 1Cs. At that point the BT probe starts rising more steeply, until it's about 25F higher than the Alt BT probe at the end of 1C and drop.

What's interesting is that if I pull the BT probe out so that it extends only 1" into the drum, like the Alt BT probe, it reads only 10F higher (preheated, empty drum). Since making that discovery, I haven't had a chance to see what difference it makes during a roast, but the effect of changing the probe depth got me wondering how batch size and the tendency for beans to move toward the faceplate interacts with probe depth. I've always thought that the higher BT probe position would be better for larger batches because they would tend to rise higher up the walls of the drum. But is it also true that they extend deeper into the drum?

I realize that BT probe temperatures are a relative measure, don't really tell you what's going on inside the bean, and they're only one tool for achieving a desired roast. But to the extent they're useful, I'm wondering whether my Alt BT probe is sufficient for all batch sizes, whether the higher BT probe would be just as useful for small batches if pulled out to the same depth, or whether the higher and deeper BT probe position is best for large batches?

Also, while I assume the higher/deeper BT probe reads higher at charge because it's further from the faceplate (closer to the depth being read by the MET probe), I'm wondering why its temperature rises more steeply than the shallower Alt BT probe just before and during 1C. Is a greater percentage of exothermic energy being released deeper in the drum? Is the expansion of the beans causing more surface area to contact the fatter probe?

User avatar
EddyQ
Posts: 1035
Joined: 8 years ago

#2: Post by EddyQ »

The Alt probe being 1/16" is going to respond quicker than the other 1/8" probe simply because its thermal mass is much less. So it will show a quicker turn point. Around 1C, your air is likely different for one probe than the other.

On my Quest M3s, I have a drilled screw on the sight glass for a 1/16" probe. I have the probe, but never used it because I didn't like flexing the cables every time I opened the door. SO, I use the higher position with a 1/16" probe. I think it goes in 2" and is close to the paddles. I've roasted down to 100gm batches with good readings on this probe so I haven't pursued using the lower possition.

As for temps within the drum when no beans are present. They are really all over the place. And airflow and temp dependent. Weird turbulence must be happening inside there. I modified my fan pot such that I can shut off the fan and I do while settling in on a charge temp. My ET probe at top, BT probe and stock analog gage all read quite differently after a long 20min soak with no airflow. Not 80 degrees tho. Maybe 30. SO, what is your air doing for your measurments?? If there is flowing air that certainly could up the deltas.
LMWDP #671

User avatar
Peppersass (original poster)
Posts: 3690
Joined: 15 years ago

#3: Post by Peppersass (original poster) »

My fan is controlled by an Artisan/Arduino/TC4/ZCD, so it can be set anywhere from completely off to 100%. My empty roaster observations were made with the fan at 30%. The discrepancy between the BT probes in 1C has happened in every roast at a variety of fan settings from 25% to 100%. It's pretty consistent, though the difference probably ranges from 20F-25F.

I agree it's not good that the lower probe flexes when the door is opened and closed. In fact, I had to replace my lower 1/16" probe not long after getting the roaster, which had two previous owners, probably due to repeated flexing causing the probe shaft to rotate in the crimp fitting at the top of the sleeve. That's one reason why I'm experimenting with the upper BT probe.

So... what's your typical range of charge temperatures? If I charge based on the lower probe, the BT might be 350F, but if I base it on the upper probe it'll be more like 430F. Drop temp will be 20F-25F higher, too.

Interesting that your upper 1/16" probe works with 100g batches. With the probe inserted 2" have you had any issues with the pounding from large batches bending the probe? I was figuring that would be less likely if the probe only protruded 1" into the roaster, but it sounds like it's not an issue.

How did you mount a 1/16" probe at the upper BT position? The hole in the larger McMaster-Carr vented screws erics recommends is too large to hold a 1/16" probe in place.

User avatar
AssafL
Posts: 2588
Joined: 14 years ago

#4: Post by AssafL »

I bent my 1/8 probe downwards. So it rests a bit more in the mass. A thought I had was to use an exposed junction TC for the beans. But it may create noise so I have not tried it.

I made the bend rather rounded so with enough coercion it is possible to insert it and remove it (but only with Eric's adapter).

Also - never bend a TC at the very tip. It may break the junction weld.
Scraping away (slowly) at the tyranny of biases and dogma.

User avatar
EddyQ
Posts: 1035
Joined: 8 years ago

#5: Post by EddyQ »

The plot below I put together a while ago to help me adjust the charge temp for different bean amounts. I typically look at roast profiles at the time the BT reaches 200F (after TP and when probe is likely close to matching bean temp). This point in time is directly influenced by charge temp and bean mass. I have two curves. The slow one works well for low grown, less dense Brazil beans and arrives at 200F in 2-2.5min. While the fast curve is best for dense beans where you arrive at 200F in 1-1.5 minutes. The plot is approximate, but does an amazingly good job with my Quest.



As for my BT probe. I have the probe offered by Shrub which comes with a pipe thread adaptor and fits nicely. No, I have not had the probe bent by large bean masses. I do have it bent slightly (like 3/8") downward, which I forgot to mention.
LMWDP #671

User avatar
erics
Supporter ★
Posts: 6302
Joined: 19 years ago

#6: Post by erics »

How did you mount a 1/16" probe at the upper BT position? The hole in the larger McMaster-Carr vented screws erics recommends is too large to hold a 1/16" probe in place.
Dick -

I don't have any of those vented screws handy but it may be possible to drill and tap the M8 to fit the vented M4.

Also, I see a machined plexiglass front cover in your future. :)
Skål,

Eric S.
http://users.rcn.com/erics/
E-mail: erics at rcn dot com

elviscaprice
Posts: 64
Joined: 6 years ago

#7: Post by elviscaprice »

Ed is using a completely different roaster with the M3s than yours Dick. I have the older style thicker drum, no back plate, modifications are drum painted black, holes drilled in one row of circumference of the rear, insulated body with exhaust wrap. I also find a greater MET/BT disparity in temperatures, anywhere from 50 to 100 degrees F, all depending on momentum, stabilized, about 60F.

I think your going to have to just learn what your roaster is capable of with the set placement of probes and go from there. Using others roasts as guidelines needing extrapolation to your model design. Once you've determined your conduction parameters from ET to BT, then determine your convection parameters with fan/airflow and how your roaster responds.

I got a very nice roast the other night for Espresso, starting to figure out my roaster. In the end it's all going to be determined by what ends up in the cup.

User avatar
another_jim
Team HB
Posts: 13871
Joined: 19 years ago

#8: Post by another_jim »

Positioning the bean temperature probe is a straightforward calibration: find the spot were the first crack begins at a reading of 195C to 200C or 385F to 395F. The crack onset temperature depends on prior roast length, so this is not graven in stone, but if you get to the first crack in around 7 to 10 minutes, this is what a well placed bean probe will show.

Environmental temperature is more arbitrary, since there are no obvious benchmarks for calibration. You'll get a lot more tipping if the beans experience environmental temperatures above 275C/525F as the first crack ends or later. If you get a tipped roast and read a lower temperature, you might want to reposition the sensor closer to the heaters.
Jim Schulman