Modding the behmor - Page 3

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
Posts: 136
Joined: 16 years ago

#21: Post by Frost »

Nice work picking apart the Behmor John, and the bean temp probe Ed! I have a couple questions... I thought the Behmor had a 'proportional' control of the heaters (like 70%, etc) so why would they shut off completely when reaching setpoint? Why not drop power to maintain? (huh, I guess the question is really for Joe to answer) Also, then, if the heaters are kept on full bore through until first crack what would environmental and bean temp profiles look like? How long does it take to reach first crack with pedal to the metal? I'm just wondering to characterize the roaster. Thanks for the grat work and reporting here on your findings!!!

User avatar
Posts: 1623
Joined: 17 years ago

#22: Post by farmroast »

It's easy enough to get a probe in the roast chamber to measure environment temperature. The real question here is: "What's the best location to place the probe?"
Not sure that environmental temp will be as useful as MET as the roaster mostly realise on radiant application with a pretty easy to penetrate mesh drum(for me a good thing). I'd want a probe/bead tip as close as possible to the outside of the drum facing the elements.

Dimmers steal too much juice so I go with a variac. I'd turn it completely manual, no circuit boards. Run heating elements direct to volt meter to variac. I'd also consider adding another element. Everything else on switches with transformer for the low voltage stuff,drum motor is 16v which I stepped up to a 50-60rpm. that will gain a little convection too. MET thermocouple just outside drum facing elements. A BT thermocouple like I added. Make a bracket to remove the drum just after roasting that would then slide into a small container/bucket cooler. Better insulate with maybe koawool inbetween the walls. Cover most of door window (except a spot to view bean) with reflective metal. For most designers it's an honor to have their original become popular with hot rodders.
LMWDP #167 "with coffee we create with wine we celebrate"

Posts: 6
Joined: 15 years ago

#23: Post by coffeecoder »

Frost wrote:How long does it take to reach first crack with pedal to the metal? I'm just wondering to characterize the roaster.
Typically, I hit first around 12 mins for 250g at 95% power most of the way. Here's a somewhat representative curve... the red curve is ET, the brown curve is bean mass temp when I use the Hottop and just another ET probe on the Behmor (located directly under the drum):

full size pic

farm, you're absolutely right about MET vs. ET. What I've been referring to as the 'ET' probe could probably more accurately be called the MET probe due to it's location. I've also wondered if, and to what extent, surface reflectivity differences between the beans and the probe tip impact the usefulness of measuring MET in a primarily radiant heat situation like this one... but I haven't taken it beyond the wondering stage.

Posts: 136
Joined: 16 years ago

#24: Post by Frost »

Thanks for posting the data John. Now is that with the available turbo-charger? I think I agree with Ed that it could use a bit more power. I would hope the heaters don't normally shut off on the way to first crack. I realise the radiant heat can be more leasure and I tend to drive the poppery a bit like a maniac, but still.....

Posts: 269
Joined: 17 years ago

#25: Post by djmonkeyhater »

Another angle for modding - I prefer to do back to back roasts with the Behmor and my roast times can grow substantially with each additional roast cycle although not consistently. My assumption is that some part of the machine is heating up and not cooling enough between the cycles. This leads to tripping of thermal protection circuits and shorter cycles of the heater elements.

And yes, there are no thermometer devices in there yet - I know, I know, I know that's bad.

I took off the side panel and found that there are 4 thermocouples (I think) in the machine.

1) under the keypad
2) in the exit path of the afterburner
3) in the exit path of the fan venting the circuit board
4) on the side of the roasting chamber under the fan.

Would anybody hazard a guess as to which one is likely the culprit? My guess is #3 but not for much of a reason. I run the fan for at least 8-10 minutes between roasts and the walls of the roast chamber are cool to the touch so I'm thinking it's not #4. My next step is to yank a plug from the board and see if it runs to try and isolate the culprit. If it won't run, I'll jumper it or just pull the probe out and let it dangle in some airstream. I think I may also run it without the side cover on for some experiments.

Another interesting note - there is another terminal for a computer power supply fan like the one that you can see in the picture. (It's near the bottom-center, to the left of the black capacitor.) If I start disconnecting the thermocouples, I have a fan I can plug in to push more air across the circuit boards. I did look at trying to remotely locate the IC board but that's a lot of soldering.

Posts: 269
Joined: 17 years ago

#26: Post by djmonkeyhater »

One more note:

I use the machine in a hood that is really effective at removing 99% of the smoke/smell related to roasting. (Otherwise the house stinks.) Here are some visuals for it. I do have some concern about it retaining some heat in the system that may be contributing to some of the lengthened roasts. I may poke some holes in it for some outside air supply but it would need to be pressurized to keep the smoke/smell in.

This was the proof of concept but the open window would allow too much smoke back in.

More evolved and substantial.

This is what is up inside. A bit noisy but it was only $18 at the mega home store.

Posts: 269
Joined: 17 years ago

#27: Post by djmonkeyhater »

more on the Behmor house.

PVC dryer pipe and the plywood to keep the smoke out. it's still in the concept phase and I'm looking to go to a dryer vent outlet and better looking wood.

fan from the bottom, residue is from 60+ roasts. It will pull almost all of the smoke outside even if I set the roast on fire. I take 80% percent of the roasts 30-40 seconds into 2nd. The chaff does go up in flames about 20% of the time. It doesn't seem to hurt much and if I'm paying attention, I rip out the chaff collector mid-inferno.

Here's the exhaust as it ships.

Taking a lesson from race cars, i'm looking to limit the restrictions in the exhaust. No extra horsepower but faster cooling and/or less backpressure/fan load can't hurt. I may shroud it upward because the cold air intake is below it. I guess that's only important for cooling.

Here's the cold air fan. I'm keeping it unshrouded on the less restriction principle although my fan/air flow experience tells me that it may need something to maintain efficiency. No idea how I'll check on that but it makes me look smart to remark that it's on my to do list.

Since it lives in the house, I don't have to worry about the cat getting caught in a thermocouple or a kid in the scroll fan with all of the panels off. Maybe a baby raccoon could climb in the hose but he'd get stopped at the bathroom fan.

Anyway, I'm wondering if disconnecting the afterburner will free up some watts in that 110v supply since I can cope with the smoke. Or at least I think I can.

Posts: 6
Joined: 15 years ago

#28: Post by coffeecoder »

I'll take a stab at your question. I've noticed that the chamber wall is still warm to the touch after a complete cooling cycle (qualifier: depending on ambient temp in the garage). Given that the main temp sensor (#4) measures temp at the chamber wall and given the 360f-400f cycle band of the Behmor's heat application, it's possible that residual heat in the wall is causing the heat cycles (or at least the initial one) to be shortened enough to notice a different in total roast time. What kind of time difference are you seeing? 30 secs? 3 mins? I could be completely off base, just my initial thought... of course, it would be an easy theory to test. Point a fan at the right chamber wall during cooldown to accelerate cooling of the area around the sensor.

Nice venting setup!

Posts: 269
Joined: 17 years ago

#29: Post by djmonkeyhater »

I'm on a business trip and don't have my roasting book with me but here's what I can remember:

My standard roast is as follows:
- 280g load
- Rosebud Espresso Blend from Atlas Coffee
- 1lb, P1, A settings
- 1st crack start on the cold machine is at the 5:00 minute remaining mark or 13:00 in.
- 2nd crack is usually at 2:40/2:30. I go 35 seconds into 2nd.
- I can tell nothing virtually nothing visually, it's all by sound.
- I hit cool and crack the door open 1" at the top. My hood will capture 99% of the smoke. After 1:00, I open the door all of the way. At 3:00 of cooling, I kill it and pull the roast. It can be handled without gloves.
- Bean weight out varies little if I manage the roast well for a net of 226-229g. 16% moisture loss or so.
- I will turn the fan back on and let it run while I get the next batch ready, pet the cat, write things down and drink some beer. Maybe another 3 minutes of cooling.
- Repeat

On the subsequent loads, I can get 1-2 minute stretches in the roast time. I really need my book for the detail but at the end of 4 consecutive batches, I can almost 3 minutes longer. It makes any sort of set it and forget totally impossible.

Here's that additional fan port on the circuit board.

Team HB
Posts: 5475
Joined: 16 years ago

#30: Post by ira »

The draw fan seems to be activated at +7:30 and stays on for the remainder of the roast. It's been a while since I dealt with the stock logic and I don't have my [few] notes handy, so I can't recall if the 7:30 is a fixed time or a percentage of the total time.
I'm guessing that the only purpose of the draw fan is smoke supression and that both the draw fan and afterburner come on at the same time. What you've described clears up a lot of questions I've had about the working of the machine.

I'd also guess the cycling pattern of the heating elements has a fair amout to do with the mechanical relay that controls them as it minimizes the number of switching cycles.