Homeroasters - Home Roasting Coffee Community » THE ART OF ROASTING COFFEE » Roasting Profiles
Who is here? 1 guest(s)
 Print Thread
Newbie, looking for advice on how I'm going wrong with my hot-rodded popcorn popper?
AMRoberts
Top  #1 Print Post
Posted on 04/23/2018 6:50 PM
I'll try and keep the background tight. I picked up the green coffee sampler and Nostalgia APH200 popcorn popper (slots around the bottom circumference of the popping chamber) package from Sweet Maria's. Did several test roasts with the popper "stock" using the Columbia Caicedo Don Juan greens, very long times that didn't seem to get to a satisfactory roast level.

So I grabbed the electrical parts junk box, ordered a couple of things I didn't have, and hacked the popper. Pretty standard pattern of mods. I'm going to skip the details since they belong on a different forum (if it becomes important I'll hang my notes into the popcorn popper roaster forum); the bottom line is I have separate manual PWM fan speed control from 18VDC (the sticker rating of the fan motor) down to a practical lower limit of about 14VDC where bean motion in the roasting chamber tends to stop, and duty-cycle control of the heat from 20%-100%.

Measurement is a four-channel TC thermometer, with bare-wire type Ks. I've got one measuring ambient, one strapped to a support strut (piece of 12ga copper wire) positioned against the outer wall so that the bead is directly in front of an inlet air slot (I call that ET), and the third on another support strut, somewhat offset from the center of the roasting chamber diameter, with the bead as high up as I could manage while keeping it below the top of the bean mass at start of roast, oriented so that the support strut blocks direct impact by rotating beans on the thermocouple bead, they have to slide around the strut (I call that BT).

With the fan at 18V, I can get at least 75g of greens into good motion before heat is applied (possibly more, I've been testing with 75g batches to learn). As the beans lose weight, and especially after volume expansion, I can back the fan speed down (and indeed must do so to get the ET and BT to continue climbing).

A big problem is that this rig is loud, as in 83-85dB, has me putting on earplugs loud, when the fan is at 18V; and not a lot better as PWM cranks the average voltage down. Overall volume trend is down, but it hits several resonant points with louder and bad tones. I have heard at least some of the first cracks, but cannot claim to have heard the span of FC as an event (start, builds up, diminishes).

I've been trying to operate through a "4-4-4" profile based on reading here and at home-barista. My interpretation of what I've read is:

1. Get the beans in motion, and keep them moving

2. Reach BT=200F within 1 minute of heat on, or 2 at most.

3. Reach BT=300F at 4 minutes

4. Ramp to BT=390F within 4 more minutes (by the 8 minute mark from heat on).

5. Keep enough heat on to reach end-of-roast target BT within 4 more minutes (by 12 minutes from heat on).

6. Heat off, run fan back up, cool beans.

I've got hand-recorded data which I'll be happy to post or send via other means if anyone would like to see it and offer advice (I'm trying to avoid an epic-length post). In descriptive terms ...

After a couple of batches, I've learned what kind of power starting point and ramp I need to hit BT=200F by at one minute, and BT=300F at four minutes. However in three batches of the Columbian run on 4/19, I was always under my target at the 8 minute mark. My last batch (with my most aggressive heater power increases) was 20F below my 8 minute target, and took 12 minutes to reach BT=433F. I had been at 100% heater power since the 7 minute mark and couldn't have dropped the fan speed much further without motion stall by the time I got there. The color chart from Sweet Maria's looked like a bit darker than their City+, not yet to their Full City shade.

FWIW the cracks that I could hear were occurring in the BT=396F - 405F range, so it seems likely that I got through FC. My ET was at least 50F above BT (476F at 11 minutes, 488F at 12 minutes), so I had "headroom" (but enough)?

We drank this effort, but my wife and I are dark(er)-roast coffee fans, not third wavers. So I believe I'm in pursuit of Full City, Full City+ or even Vienna roast levels, and I don't see how I can get there on a 4-4-4 profile. It seems like I'd have to allow a much longer third phase and see if BT would continue to rise, but isn't that the road to baked coffee?

That was the last of my Columbian, so on 4/21 I tried four batches of Ethiopia Yukiro Cooperative. Things got even more frustrating. While the Columbian beans made a consistent green-to-yellow transition during the first 4 minutes, the Ethiopians were visually a mix of greenish-yellows and tans all the way out to 6-7 minutes, and took me 13-16 minutes to reach BT=435F. Overall color once again looks between City+ and Full City, but I had half dozen or so obviously under-roasted beans in each batch (I culled them out). The Ethiopian beans seem a lot more resistant to heat, is that expected?

I'm going to wrap this up, I can add data or send it via PM if anyone wants to see more, but what is the "right answer" when pursuing a darker roast ... Let the time extend beyond 4 minutes after FC, or full-throttle the heat at the start of the roast and keep reducing airflow to the edge of motion stoppage, so that I'm getting to FC earlier than 8 minutes into the roast?

Many thanks,
Alan
 
AMRoberts
Top  #2 Print Post
Posted on 04/24/2018 8:53 AM
Thanks chaff, I'm still trying to avoid stirring since I've got a thermocouple kinda-sorta in the way, but I can probably still increase batch size some and still have beans in motion from airflow, so I'll try experiments with a bigger batch and with more heat sooner.

The DC PWM controller I have supplying the fan motor is probably not a triac (haven't opened it up, but the typical 555/other to vary duty cycle, driving a power MOSFET switching DC on/off of the motor would be my guess), but it is likely to still have fast rise/fall times. I'll give the capacitor some thought.
 
renatoa
Top  #3 Print Post
Posted on 04/26/2018 3:46 AM
I find this part questionable:

"2. Reach BT=200F within 1 minute of heat on, or 2 at most."

The BT heat increase evolution is given by the following equation, pure thermodynamics:

BT = ET (1 - exp (-t/k))

where t is time and k is a factor that include the whole system thermodynamics: beans heat absorption capacity, heat losses, etc.
If you are playing with values and find a value for k to ensure an average drying temperature by minute 4, and FC in the minute 8 ballpark, you can obtain a "perfect" roast profile, as attached.

According to some industrial roasting research, yeah, I know, commercial roasters, they sucks Grin... is not good to start the roast with more than 200 C air temp for the first 2-4 minutes.
If accepting the above limiting of ET at 200 C in the first 4 minutes, according to this profile/equation, then real BT, I mean inside bean temperature evolution, will be as follows, by minute: 51-90-122-148

What people measure as 200F (93C) at one minute is simply bogus, an average between hot air and beans, this heat simply can't absorbed by a bean, at least not theoretically.

You can proof this simple, blow for two minutes with 200 C, stop the blower and let system temperature stabilize... will have a real BT measurement.
Or spill the beans in a preheat thermos fitted with a thermometer.
renatoa attached the following image:
simroast.png
Edited by renatoa on 12/10/2018 12:53 PM
Roasters: DIY: TO based IR 200-640g
Commercial, moded: Skywalker, Dieckmann RoestMeister, Nesco, popper(s).
Electronics: TC4ESP (author), MS6514, USB/Artisan/Apps (contributor), PID controllers
Grinders: Potu ghost burrs, Arco, Xeoleo ghost burrs, Krinder, vintage PeDe Dienes, MBK Feldgrind, Kinu M38
Brew/presso: ALM pour over, Arin lever, Staresso Mirage, Hario Cafeor dripper, Flair Signature, Kompresso, Aeropress, Gaggia Mini
 
ThomasCee
Top  #4 Print Post
Posted on 04/27/2018 4:46 PM
Hmmmmm..... You are soooo much more detailed in your popper roaster setup than I am. I have a simple Variac that only controls input voltage; so the heat and fan go up and down simultaneously. The clumsy-ness of it drives me nuts. Simply awful. My only monitoring is my senses: smell, sight/bean color, sound, and a stopwatch on my watch.

BUT...

I have roasted some really fun batches that my coffee snob family has really enjoyed.

I have noticed this: There are many people with differing setups who roast different ways and get results they like. I have seen YouTube tutorials that say differing things. Heck, I have a coffee shop near me that has a 10Kilo Proaster ( https://www.firstcrack.com/product/proaster-10kg-coffee-roaster/ ) who leave the fan at ONE speed and the heat at ONE setting and run it all the way and drop on temperature (425) at about 14ish minutes and record their roasts on EXCEL! I'm like, "$27,000 roaster and you don't use any of it's beautiful features???"

My point is, all theories and guidelines are secondary to the taste of the cup. You like darker? Simply roast longer and drop by sight/bean color, and smell. Do many roasts this way till you start getting a feel for what tastes good to you both. Record those milestones as you go so you can begin to see patters emerge.

Bottom line: Use tools to help you get a repeatable good cup; don't pursue a tool ideal and cross your fingers that you like the cup.

Just MO!
 
btreichel
Top  #5 Print Post
Posted on 04/27/2018 9:48 PM
You have to split your fan from your heater coil. best is a variac for both, but a fixed fan works.
 
AMRoberts
Top  #6 Print Post
Posted on 04/30/2018 5:14 PM

Quote

renatoa wrote:
...
What people measure as 200F (93C) at one minute is simplu bogus, an average between hot air and beans, this heat simply can't absorbed by a bean, at least not theoretically.
...


renatoa, thanks for the information. I appreciate the error potential for my thermometry. My ET probe is directly in front of one of the air inlets to the roasting chamber so I believe that is giving me a close value, but all that I'll claim for my BT probe is:

1. I did what I could to get it away from the inlet air while staying under the top of bean mass at the start of roast.

2. It certainly lags the ET during the roast, and stays above ET during the fan-only cool down.

3. When I manage to hear any of 1st crack over the fan noise, it has been in the 395-402F, across several different types of beans.

Nonetheless, I realize it is at best some blend of bean surface temperature and air temperature, and I take the point that it is likely to be greatest error from internal bean temperature early in the roast.

Cheers,
Alan
 
AMRoberts
Top  #7 Print Post
Posted on 04/30/2018 5:32 PM

Quote

ThomasCee wrote:

Hmmmmm..... You are soooo much more detailed in your popper roaster setup than I am. I have a simple Variac that only controls input voltage; so the heat and fan go up and down simultaneously.
...


Hello ThomasCee. Possibly this belongs in a roaster forum, but ... My biggest reason for doing the split control before I made the mods was the desire to let the popper do most of the cool-down of the roast "hands off."

After making the changes and putting it all back together, I found that without dropping the airflow during the roast I couldn't make it to desirable ET, even with the stock thermostat removed from the circuit.

If you can get there by dropping AC voltage to an unmodified popper, I'll speculate your popper has a higher-wattage heating element than mine has.

Quote


... You like darker? Simply roast longer and drop by sight/bean color, and smell. Do many roasts this way till you start getting a feel for what tastes good to you both. Record those milestones as you go so you can begin to see patters emerge.


Working on it, but isn't there an issue with too much "roasting longer?" I thought (from forum reading), that too long a roasting time led to dull/baked flavors.

Cheers,
Alan
 
ThomasCee
Top  #8 Print Post
Posted on 05/02/2018 1:35 PM

Quote


Quote


... You like darker? Simply roast longer and drop by sight/bean color, and smell. Do many roasts this way till you start getting a feel for what tastes good to you both. Record those milestones as you go so you can begin to see patters emerge.


Working on it, but isn't there an issue with too much "roasting longer?" I thought (from forum reading), that too long a roasting time led to dull/baked flavors.

Cheers,
Alan


Yes indeed.
I think my point is, don't be afraid to roast too dark on purpose just to see how it tastes. Each "less than perfect roast" adds to the knowledge base. I have this feeling you are working on the nuances of profiling when you need to just play with adjusting drop time for 5 or 10 or 20 roasts till you generally like the taste, THEN start nuancing all the details. You'd be amazed how basic some places roast, yet they have a taste they like and they can roast it repeatably.

The other thing I don't remember if was brought up or not, was beans. I definitely prefer some beans, and find others to be relatively simple/boring/uncomplex, whatever you want to call them. My first beans I roasted were some Brazilian Naturals, and while they were nice, it really had me confused for many many roasts. I could not get very much interest out of them no matter how I roasted them. Burnt, puckering-ly underdeveloped, nothing. I then got on the cursed "try all the beans" bandwagon and have had a blast tasting the different regions since then!! The conclusion for me is that beans make a MASSIVE difference in taste, regardless of how they were roasted.

Again, all of this is just the opinions of a fellow roaster who has been having a blast drinking his own home roasted coffee.
 
ThomasCee
Top  #9 Print Post
Posted on 05/02/2018 1:36 PM

Quote

btreichel wrote:

You have to split your fan from your heater coil. best is a variac for both, but a fixed fan works.


Yes indeed, but I'm gathering goodies to make an affordable drum roaster next smile
 
btreichel
Top  #10 Print Post
Posted on 05/02/2018 11:34 PM
What we wouldn't do for an inexpensive ( 40 to 70) NG control valve. Kills.me that the valve can be 2 to 3 times the price of the rest of the control system.
 
renatoa
Top  #11 Print Post
Posted on 05/03/2018 2:11 AM
Why not fit a RC servo on the shaft of actual manual valve ?

The servo can be controlled by an arduino usb to pwm interface.

A video showing how this works:
https://www.youtu...-YLL6bov_4

and attached a picture of of a gas valve shaft, M6 thread, to have an idea about the hardware needed to adapt a RC servo.
...
renatoa attached the following image:
ozturk_valve_shaft.jpg
Roasters: DIY: TO based IR 200-640g
Commercial, moded: Skywalker, Dieckmann RoestMeister, Nesco, popper(s).
Electronics: TC4ESP (author), MS6514, USB/Artisan/Apps (contributor), PID controllers
Grinders: Potu ghost burrs, Arco, Xeoleo ghost burrs, Krinder, vintage PeDe Dienes, MBK Feldgrind, Kinu M38
Brew/presso: ALM pour over, Arin lever, Staresso Mirage, Hario Cafeor dripper, Flair Signature, Kompresso, Aeropress, Gaggia Mini
 
btreichel
Top  #12 Print Post
Posted on 05/03/2018 8:36 PM
Actually I have a stepper motor to try that with. It's commonly used for 3d printers. From what I hear it can take multiple turns. The trick is the little drive board, they can be a bit finicky; but got that dialed in, then didn't attach it.
 
renatoa
Top  #13 Print Post
Posted on 05/04/2018 12:45 AM
Depends on your valve force required, the unit from the picture above can be moved very easy, measured 300 grams with a 10 cm handle, that equates to 3kg.cm or 0.3N.m , torque available from most standard RC servos priced under $10.
The complete range of this particular valve is 90 degrees, no need for multi-turns, but there are available such servos too, robotic or navy winch servos are designed to rotate endless.

The controller also can be bought ready to use, at arduino price, check this: https://www.polol...oduct/1350
Featuring a complete software system, including a scripting system, with programmable acceleration, speed and target position, this controller can be programmed by itself to ensure a profile following, without any help from another software.
... worth a try, so I ordered one, especially because marvelously, it is available locally. Grin
...
Roasters: DIY: TO based IR 200-640g
Commercial, moded: Skywalker, Dieckmann RoestMeister, Nesco, popper(s).
Electronics: TC4ESP (author), MS6514, USB/Artisan/Apps (contributor), PID controllers
Grinders: Potu ghost burrs, Arco, Xeoleo ghost burrs, Krinder, vintage PeDe Dienes, MBK Feldgrind, Kinu M38
Brew/presso: ALM pour over, Arin lever, Staresso Mirage, Hario Cafeor dripper, Flair Signature, Kompresso, Aeropress, Gaggia Mini
 
ThomasCee
Top  #14 Print Post
Posted on 05/04/2018 11:21 AM
Any new successful roasts you n the wife are loving AMRoberts?
 
AMRoberts
Top  #15 Print Post
Posted on 05/04/2018 2:52 PM

Quote

ThomasCee wrote:

Any new successful roasts you n the wife are loving AMRoberts?

We did rather leave the roast profile question behind for a diversion into drum roaster gas control hardware, didn't we? Roflmao

I'll admit you had me worried with:

Quote

... My first beans I roasted were some Brazilian Naturals, and while they were nice, it really had me confused for many many roasts. I could not get very much interest out of them no matter how I roasted them. ...

Since my attempts with Sweet Maria's Brazil Dry Process Santa Luzia Yellow Bourbon are what we had yesterday and this morning. These were roasted longer than I had ever gone before (15-17 minutes) to reach 454F on the thermocouple that I call BT. Better than anything else I'd done earlier, still not where I want to be in terms of taste.

Should have been obvious, but I've discovered the down side to the one pound samples from Sweet Maria's ... Not really enough of any one bean for a beginner like me to learn the bean's response to heat and vary the roast level. I'm out of the Brazil Santa Luzia, and will be switching to the next bag out of my samples box: Guatemala Xinabajul San Pedro Necta ... Undoubtedly will change things. So I'm kinda-sorta following your try different beans path. My next order of greens will be different from anything in my sampler, and 5-lbs of each so that I can experiment/screw-up in more detail for each type.

Cheers,
Alan
 
btreichel
Top  #16 Print Post
Posted on 05/04/2018 10:51 PM
I talked with some north roaster guys, they were talking over 1 turn. P.S. the Arduino can do two servos from pins 9 & 10, other pins via software servo.
 
AMRoberts
Top  #17 Print Post
Posted on 05/05/2018 10:21 AM

Quote

btreichel wrote:

What we wouldn't do for an inexpensive ( 40 to 70) NG control valve. Kills.me that the valve can be 2 to 3 times the price of the rest of the control system.

btreichel, have you looked at:

https://forum.hom...ad_id=3506

in the Electric and Gas Heat Sources forum, and:

https://forum.hom...ad_id=3453

in the Fluidbed Roasters forum?

I've only skimmed them since I'm not at that point in my roasting journey, but both threads talk about proportional gas valves (Kelly Pneumatics, Clippard) that are < $100.

Those might meet your needs, although you would need to either buy or design and build the rest of the gas train. The modern on/off gas combination valves used in the furnace and cooking appliance industries have time-tested safety features built in, remember to think about safety and faults if you are rolling your own!

Cheers,
Alan
 
ThomasCee
Top  #18 Print Post
Posted on 05/05/2018 4:28 PM

Quote

AMRoberts wrote:

[quote]ThomasCee wrote:

Any new successful roasts you n the wife are loving AMRoberts?

We did rather leave the roast profile question behind for a diversion into drum roaster gas control hardware, didn't we? Roflmao

Lol, yep, we sure did go off topic!

Interesting on the Brazilian. Sorry, didn't mean to shoot the exact beans you were using BBQ grill ! I'm not sure what taste you are going for, but your experience does indeed sound similar to mine.

The Guatemala's I purchased, I have enjoyed very much. Seemed to be a very "full bodied" cup; my brother described it as a "big tasting dinner coffee" (I have such a terrible time describing taste) and we overall liked it very much.

I agree on the Sweet Maria's sample packs, they go very fast. I have also ordered from RhoadsRoast Coffees, the sweet lady there returns every phone call and will chit chat coffee and roasting ideas for the greens.

Welcome to coffee roasting! Hope you enjoy the coffee journey as much as I have. I'll post some of my roasts here next time I roast.
 
renatoa
Top  #19 Print Post
Posted on 05/06/2018 1:39 AM
Wondering if this forum software has any feature to move some off-topic posts into a complete new thread...
This is a frequent practice in our national coffee forum, lots of derailing there Grin
 
AMRoberts
Top  #20 Print Post
Posted on 05/07/2018 10:32 AM

Quote

ThomasCee wrote:
... The Guatemala's I purchased, I have enjoyed very much. Seemed to be a very "full bodied" cup; my brother described it as a "big tasting dinner coffee" ...

Hope I can get there, I've got a Mason jar full of my attempt on the Guatemala Xinabajul San Pedro Necta next in the drinking queue.

I ended up changing more things than just the beans (bad experimental practice). I have an open-top box shielding the popper from breezes, and casual exhaust air recycling ... A leaky duct steering exhaust air to a screened port in the side of box. Chaff stops at the screen, but warmed air makes it into the box, increasing inlet air temperature for the popper.

Most importantly, I've run enough batches through this popper to realize I've got two paths for reaching a dark (FC+ / Vienna) roast:

1. If I run the 4-4 drying and ramp to first crack time, then the overall time for a dark roast is going long (16-18+ minutes). Or,

2. I crank things up harder early in the roast (shorter drying time, shorter ramp to first crack), and I can reach FC+/Vienna in 11-12 minutes.

Since my samples include two bags of Guatemala, I decided to try the second approach for my roasts this weekend. I held ET=350F for only 3 minutes, then went to 100% heater power and backed down airflow as much as possible while keeping the beans in motion. When ET made it to 480-490F, I held that temperature until ending the roast.

Start of first crack was consistent at 6 minutes, ending close to the 8 minute mark. For the first time I could actually hear the second crack. Of course I don't know if this was due to the different beans, because I was using more aggressive heat, or because I had ear protection on until I reached lower fan speeds. End of roast was 11 minutes for the two batches I'd call Full City+, and 12 minutes for the three batches I'd call between FC+ and Vienna.

I guess I'm risking underdevelopment, but the sample beans from each batch that I cut open have consistent interior color.

Quote

... I'll post some of my roasts here next time I roast.

Looking forward to that, thanks!
 
Jump to Forum: