KapoK gas roaster

[hankua]

Here's a link to a KapoK 1.0 Youtube video

http://youtu.be/aObo0DvCD7A

[Cammie]

Congratulations!  That is a beautiful roaster.

[John F]

Slick!! 

[9Sbeans]

Color coded buttons –

Green
The green button (pic-21) will light on as long as there is juice.  Press this green button to power up the system, including the drum motor, vent fan, digital BT & ET displays, and the white LED lamp.

Yellow
After the system powers up, the gas sensor will detect the pressure at the inlet.  (pic-22) Turn on the gas valve of the tank, with the gas inlet pressure, the yellow gas indicator lights on. 

Red
(pic-23) Press the red ignite switch button, fire up.  Press it again to turn off the burner; red button light off.

[9Sbeans]

White
(pic-24) Press the white button to turn on the fan of the cooling tray.  Press it again to turn it off.

Blue
(pic-25) This button is at the bottom of the control panel.  Press it once to activate, and yes, it’s Bluetooth ready.  It can also be wired to a computer via the connector, as can be seen in a test run picture taken by the manufacture (pic-26).       


I was playing with the gas-flame control dial these days and haven’t got into the software (Artisan) yet...   

[9Sbeans]

Vent speed

(pic-27) The air flow is manually controlled by the vent speed knob.  Manufacture suggests 3.0 as a starting value and 5.5 for a fast system cool down after roast.  Without the flame, I could dial the fan speed from 0 to 8.5.  A user’s review said that the DC motor of the fan is regulated by an IC-controller for voltage regulation , and he has measured the air flow rate being between 0.1 to 1.2 m^3/min.   

I tried vent speed from 3.0 to 2.5 only once while burner is on, and observed the increased temperature rising rate.  Conversely, vent speed 3.5 and 4.0 brings in cold air and slowdowns the temperature rising rate (note that the BT is still rising, just rises slower). 

The vent provides slightly negative pressure inside the drum, and the negative pressure is monitored by the pressure sensor as shown in pic-16.  An anecdote: A K500 user said that he once improperly installed the exhausting hose, and gut wind blew the air backward from the exhausting pipe to the roaster.  The gas line was therefore shut off, leaving that user puzzled by the “igniting device malfunction”.  Theoretically, if I drop the vent speed low enough, I should be able to test the safety gas shut-off function as advertised.  However, I don’t want to test this safety feature, since insufficient vent is usually associated with dangerous accumulation of carbon monoxide.   Unconfirmed feature for now.

[9Sbeans]

Automatic BT Mode
The temperatures dial (pic-27, lower right, currently set to 50 degree C) controls the switch between manual and auto BT mode.  IMO, the auto mode is valuable in the warming up of the system and in the maintaining of charge temperature. 

Manufacture recommends slow heat up steps (for the lifetime of the drum) and 15-mins of equilibration time at charge temperature for the first roast of the day.  Dialing this temperature control knob, I can set the target BT temp at 50 degree C, and gradually increase it  to 180 degree C (the charge temperature) and would never worry about overheating of the drum.   In the auto mode, the gas pressure is switched between “High, Low, and Off” stages.  For example, if the current BT is at 40 degree C and target BT is at 120 degree C, the gas will be in “High” until close to 117 degree C, and then switch to “Low” for slower approach, or “Off” for an overshoot (manufacture states 3.5% error of auto BT).

I actually roasted half-pound of green beans by the auto mode, and it tasted really good.  I charged the beans, simply dialed the target BT at different time points, and voila I got Guatemala Antigua - Finca Pavon in Full City Roast.  But, no thanks, not again, I’ll do the full manual roast. 

To do the manual roast, just set the BT dial to a high enough temperature (250 degree C), and the system will be always in the “High” gas pressure mode.  The user thus controls the gas pressure gauge (as shown in pic-10) and the vent speed, the same as all other gas roasters.

[9Sbeans]

Those look like ribbon burners, do they change from two row to three by gas pressure adjustment? What's the maximum Kpa possible? 3Kpa is the max on the Mini500+.

Hank, I think you are right.
2 rows of flame could be due to the low gas pressure.

***

As shown in pic-10, the gas pressure gauge can take up to 600 mmAq pressure.  I tested the max pressure output again this afternoon, and confirmed that it could easily reach 550 mmAq (I didn’t go higher; last time I stopped at 350) with Steve’s regulator assembly. 

The “High” pressure of the auto-mode is manually set by the gas pressure gauge.  The “Low” pressure of the auto-mode is a proportion of the “High” pressure, but I don’t know the formula.   For example, if I manually set the “High” to 280 as manufacture suggested, the “Low” pressure was about 38 mmAq.  However, when I tricked the system and set “High” to 80, the “Low” was about 25mmAq. 

[9Sbeans]

Noise
The roaster is made of stainless steel and cast iron, and the thermal expansion coefficients of these materials are different.  Cold drum makes loud noise when rotating, and the noise gradually disappears when BT rises above 75 degree Celsius.  The user’s manual indicates that the drum has been aligned when BT is at 200 degree Celsius.  The fan of the cyclone is very quite.  The fan of the cooling tray is a little bit louder than that of the cyclone. None of those noises would mask the sound of 1st or 2nd cracks.   

[9Sbeans]

Bluetooth/wired Artisan support

It was effortless to set up the Bluetooth connection.  In the Artisan software, BT & ET were configured as “Delta DTA PID temperature controller”. 
I was not familiar with the software, and anyhow got my first roast with the data logging.  As can be seen in the plot, before charge (-1 to 0 min), I set the BT targeting at 185*C to warm up the machine, and the BT was hovering around my target temperature.  After charge, the turn point was around 70 ~ 80 seconds (shown 1:09 on this plot).  I increased the power (propane pressure) at around 8:30 (@ BT 179*C), and increased ventilation (from 3.0 to 3.5) at the 1st crack (BT 191*C).  I was playing with vent speed (increasing 0.1 every 15 sec until 4.5), and it seemed that the increasing of vent speed brought in cold air, and it decreased RoR while maintaining RoR at positive range.  The ET was lower than BT after 1st Crack; nevertheless, BT carried on increasing and did not stall.  I did not want to reach the 2nd crack, therefore didn’t know exactly the BT readouts at 2nd Crack.  The stretching of development (3:50) might be too long, and I will know the results 4 days later. 
I will try to work out at least two different profiles, reproduce them consistently on the same types of beans, and then compare tastes in the cup.  KapoK and its users have accumulated hundreds of roasting profiles, and IMO this database is invaluable.  In summary, the machine and software were functional and well integrated.  The sky’s the limit, and the only questionable link is now the person operating the machine.    

[9Sbeans]

Two more batches today, I tried faster roasting time and declining RoR.

01-Ethiopia Dry Process Gr. 1 Yirga Cheffe Kochore
Weight loss 15.0%
Originally I wanted to test the 2nd crack (still don’t know my BT readouts @ 2C), but chicken out again.  Personally I am not a fan of dark roast.  Just don’t want to “waste” the beans.  Probably the next time. 

02-Ethiopia Wet Process Gedeo Yirga Cheffe
Weight loss 13.9%

Ps. I shrink the size of the plots; didn’t know how to do that yesterday.   

[hankua]

Nice!

What charge weights are you using?
I use 180c for charging 1lb. You might want to try charging @150?

[9Sbeans]

 

Nice!

What charge weights are you using?
I use 180c for charging 1lb. You might want to try charging @150?

I use 227g (8oz). 

Lower charging temperature is usually suggested for natural/dry process beans and for smaller charge weights.  For the same Dry processed Ethiopia, 8oz, I actually had tried a charging temperature of 200*C (392*F) to test the higher boundary, and apparently the beans were not scorched at all.  I think the dual-layered solid drum in K500 had prevented beans from scorching of the direct-fire.  We may see different responses in a perforated drum (as in a Mini500+ or a Huky500) with the same CT. 

Surely will also do a 150*C charging temperature and report back.    Might be couple weeks later though; got enough roasted beans for now.

[9Sbeans]

Quick follow-ups of previous batches:

The 20150110-test1 was Espresso Workshop #30 - Harmonic Tremor from Sweet Maria’s.  It was a blend of Rwanda, Guatemala, and El Salvador. 
Charge Mass: 227g
Charge Temp: 364.5F
Dry/Ramp/Development: 4:04/6:12/3:50
FC-start temp: 374.7F
Finish Temp: 400.1F
Overall Roast Time: 14:06
Moisture Loss: 14.5%
I stretched the ramp and played with the ventilation speed after the 1st crack.  Day 4 & day 5 post-roast, it was sweet.  However, on day 6, the wife did not like our morning espresso.  It was still sweet and there was long “baked chocolate” after taste, but it has no acidity, lost flora notes.   

I blended the Espresso Workshop #30 with the DP Ethiopia roasted on 01/11.  1/3 of #30; 2/3 of DP Ethiopia for the wife; 1/2 of each for myself.  Fantastic.

The WP Ethiopia was in my office as my after lunch coffee, with Lido2 & Aeropress.  The dry fragrance was phenomenal on day 3, and it was a vivid, sweet cup since. 

In short, I will stay with Scott Rao’s suggested profile as my standard profile for now, and experiment different profiles.  It is very easy to reproduce the S curve on the KapoK 500 sample roaster.  I’m aware of the another_jim/Rao’s debate, hence will work on SlowStart/FastFinish profile in next couple months.

[9Sbeans]

I did two more batches yesterday; with high (363.5*F) and low (303.4*F) charge temperatures.

Roasting Info:

Bean: Kenya Nyeri Kagumo Peaberry - Wet Processed
Roaster: KapoK 500 sample roaster
Charge Mass: 227 g
Charge Temp: 362.5F / 303.4F
Dry/Ramp/Development: 3:34/3:33/2:30; 6:30/3:30/2:18
FC-start temp: 374.2F; 374.2F
Finish Temp: 400.8F; 400.8F
Overall Roast Time: 9:38; 12:18
Moisture Loss: 14.1%; 14.1%

Unfortunately, the Artisan crashed when I wanted to save the low charge temperature profiles, and the Dry/Ramp/Development times were my recollections.  I kept most parameters identical for both batches, and the roasting profiles of the low charge temp were very similar to that of the R&LT results from Mark ([creative nickname]) and Rich (boar_d_laze) who have 1# USRC sample roaster.  (With the 227g charge @ 300F, my TP was about 200F)

I have attached the roasting profiles of the high charge temperature.  Temperature in Fahrenheit.  Both Power (P) and Fan (F) are in %.  Assuming 500 mmAq = 100%, the pressure gauge knob allows me to dial it in 5-mmAq increment, which means 100 steps of control in Power.  The fan speed in theory is from 0 to 8.5, in 0.1 increment, with useful range 3.0 to 5.5 (without stalling BT), 25 steps of control in fan speed.

When I feel ready, I will join the R&LT in HB, and will not update this thread.  Cheers all.