HomeRoast Digest

Topic: Not so stock PII versus FR (17 msgs / 506 lines)
1) From: Jim Schulman
Hi All,
I've about given up on commercially produced home roasters, and 
all I'm looking for is anything that blows enough hot air to 
hook up to my Variacs and TC. So I modded a PII for separate fan 
and heat control and ran it next to my similarly modded FR for a 
few comparisons. I fitted the PII with a can extender that 
allowed me to use the FR lid as a chaff collector. It roasts 
about 140 grams as readily as the FR. 
There is only one obvious physical differences in the roast 
process: bean motion. In the PII the beans spin horizontally 
like in a Rosto, in the tilted FR, they circle vertically and 
pulse up and down.
The PII is virtually as easy as to the use as the FR, except I 
have to unplug to empty the beans, rahter than just switching 
roast chambers. Given the cost difference, the PII is a bargain, 
even if one buys the separate FR lid as a convenient chaff 
Profiling is marginally more difficult on the PII, since the 
heavy aluminum roast chamber increases the lag between heat 
changes and bean temperature response.
The first roast on the PII was a dud for the comparison, since 
it turn out my TC on it reads 10F lower than the FR's for the 
same degree of roast. But the next two pair were OK with 
identical pofiles and no obvious color difference or crack stage 
in the roasts. This was rather good news, since I'm roasting 
without seeing the beans on the PII.
The taste test showed one significant difference between the 
roasters. On the trivial side, I scored the FR roasts 5% higher 
on aroma and flavor, the PII 5% higher on body and sweetness. 
This is consistent with a remaining slight discrepency between 
degree of roast too subtle to show on visual inspection. But 
finish was a surprise, the PII roasts were consistenty smoother, 
with a complete absence of the slight prickliness I often get 
with FR roasts. This was especially clear on espresso shots. 
What makes me curious is that this is even true of the way too 
dark PII roast, which should have been much bitterer in the 
This is early news, so I don't know if these results will be 
consistent or not. If they do hold up, it means that there are 
significant taste differences 
between roasters that do not derive from the profile, but from 
the particular 
ways the heat is transmitted to the beans.

2) From: Ken Mary
Do you mean PI?
I suspect a difference between an Aromaroast with a full screen bottom and a
popper with the aluminum bottom. In the very few comparisons that I have
done, the Aromaroast gave better "roast" flavors than the popper. The screen
bottom should transfer little heat to the beans while the popper's aluminum
should transfer a significant part of heat by conduction.
In lighter roasts short of second, these roast flavors are not significant,
so either roaster can produce good results.

3) From: Jim Schulman
On 7 Nov 2003 at 7:17, Ken Mary wrote:
Haven't stumbled across the PI yet; the PII is flimsy in most 
places, but the roast chamber is a fairly thick aluminum; and 
certainly has more of a thermal mass than the FR's glass roast 
chamber. The other popper I tried, can't remember the make, had 
very flimsy metal, like a cheap flue duct, thinner than a tin 
can. That one started walking around the counter and howling 
when I turned up the fan to move the beans.

4) From: Felix Dial
Hi Jim, welcome to popper roasting.  you wrote:
The beans in both my P2 and 1250watt pumper (pumper2) also spin around
horizontally in the beginning.  But as the roast progress, the beans still
go around and around, but most of the motion is up and down.  Have you tried
removing the FR lid late into the roast to see how the beans move?  When
roasting big loads with a p1 or 1400watt pumper, I often tilt the popper by
wedging the spoon end of a wooden spoon under the roaster.  The beans in
this case move like ocean waves cresting on top of each other instead of
spinning around and continue moving in this way till the roast ends.
Towards the end of the roast, if I remove the spoon making the popper level,
the beans move up and down en masse.  The beans kind of float around, almost
hover, and then they plop up and mix around when air needs to escape.
IIRC, in your FR, you're thermocouple is placed just above the moving beans?
Is the thermocouple in the P2 placed in roughly the same position?  I don't
know if this might make a difference.
Do you plan to conduct more tests with other types of poppers?  Please keep
us posted.

5) From: Jim Schulman
On 7 Nov 2003 at 21:57, Felix Dial wrote:
I didn't know that, thanks for the tip. I'll tilt the roaster 
next time.
Roughly same height aboove the inlet in both cases. I'll try to 
adjust the PII position to get the same readings for the same 
roast degree as the FR, since that'll save me conversion 
Stay tuned,

6) From: Rich Adams

7) From: Ben Treichel
With my P1 with a glass top, right into the metal base, I notice a swirl 
burble action. The beans burble up on side, and as they travel down the 
next side, at the metal base that starts to pick up some side/swirl motion.
I speculate that they travel 1/2 a turn, before the burble up again.
FYI, I tilt the back up about 3/4 inch. Have since curmudgeon, non 
purchasable virtue, far south Texas wandering John, suggested it a year 
plus ago. (wanted to be clear in the id ;-) )
Rich Adams wrote:

8) From: Felix Dial
Hi Jim, you wrote:
... snip ...
My first roast of the Organic PNG that Tom ran out a few weeks ago I took in
my poppery2 the first few snaps of 2nd, 445° in my setup.  The resulting
coffee was excellent.  All the flavors that Tom spoke about in his review
were all there, pears, hints of chocolate, just wonderful.
The next roast, I used my 1400W Pumper and increased the roast batch from
3.5oz to about 6oz.  I followed the same profile, the same final
temperature, and the same few snaps of 2nd crack.  Unfortunately, this roast
wasn't as good as the first.  The great flavors were there, but very muted.
This morning I just finished drinking the very last of the one pound ORG PNG
that I purchased.  I roasted the remaining 6.5oz in a poppery1.  Again same
profile, same final temp, and same 2nd crack as the very first roast.  The
results were better than the roast from the big pumper, but no where as good
as the very first roast.
Any deviations in flavor I've just attributed to slight deviations in
ramping, and not the poppers themselves.  I use dimmers and have to twiddle
them manually - no computer controllers here.  Sadly, Tom is all out of the
Org PNG, so no more testing with this bean.
Do you have one of the 1250Watt WearEver Pumpers?  See url below for a
photo.http://members.cox.net/felixdial/WEP2s.jpgI don't know how much testing you plan on doing, but I'd be willing to
contribute to your effort by giving you one of my spare pumpers.  Just shoot
me an email off list with your address if you like.

9) From: Jim Schulman
Hi Felix, you wrote:
I don't think the very small profile differences you'd get 
because of this would be the cause. I'm going to focusing on two 
things, one of which I think is a red herring and the other 
- 1. Air/beans ratio: some say that lots of flowing air will 
strip off the volatiles more readily than stiller air. Given the 
very high pressures within the bean during roasting, I think 
this is a red herring, since the extra airflow won't add much to 
the pressure induced tendency of the beans to outgas. However, 
I'm often wrong.
- 2. Supply air heat versus bean weight. It could be if the 
supply temperature is much higher than the bean temperature, and 
the bean mass is too small to absorb the supplied heat, it will 
damage the beans and taste. It may be this, rather than the 
profile itself, that improves the taste of slowly ramped air 
roasts (everyone, except Ken the iconoclast, agrees on ramped 
supply air temperature roasts tasting better than constant 
temperature ones). It could be that the WEP, PI, and PII require 
very different supply air temperatures to achieve the same 
I'm not sure how to test these ideas. Simply overloading an air 
roasters, so it requires less heat and air per bean, doesn't 
seem to produce impoved roasts. This seems true even when the 
temperature stays in control and the roast remains even. The  
heat absorption capacity of the beans may be compromised by the 
reduced heatloss.
Thanks, Felix, I'll do that. But it'll take me a couple of weeks 
to get thought out and organized for the experiment.

10) From: Ken Mary
From dictionary.com:
Word History: An iconoclast can be unpleasant company, but at least the
modern iconoclast only attacks such things as ideas and institutions. The
original iconoclasts destroyed countless works of art. Eikonoklasts, the
ancestor of our word, was first formed in Medieval Greek from the elements
eikn, ³image, likeness,² and -klasts, ³breaker,² from kln, ³to break.² The
images referred to by the word are religious images, which were the subject
of controversy among Christians of the Byzantine Empire in the 8th and 9th
centuries, when iconoclasm was at its height. In addition to destroying many
sculptures and paintings, those opposed to images attempted to have them
barred from display and veneration. During the Protestant Reformation images
in churches were again felt to be idolatrous and were once more banned and
destroyed. It is around this time that iconoclast, the descendant of the
Greek word, is first recorded in English (1641), with reference to the
Byzantine iconoclasts. In the 19th century iconoclast took on the secular
sense that it has today, as in ³Kant was the great iconoclast² (James
I do not seek to overthrow, but to question all things. I started with
ramped roasting, a necessity for the Melitta Aromaroast. That was carried
forward to poppers, where I began to see that faster profiles yielded better
flavors. So of course, the question "How fast?". That quest ended
temporarily at about 2.5 minutes. But now to investigate ramping again, I
must slow the profile to enable manual control. Presently I am finding 5 to
9 minute roasts near the optimum, depending on the equipment.
Not the supply temperature but the total energy input. Drum roasters
instantly heat the bean surface in contact with the drum. Some air roasters
heat the beans through mostly convective transfer (Sivetz and Melitta).
Other air roasters transfer a large part of the heat through conduction and
radiation from the hot metal chamber bottom. The optimum profiles and final
results may be widely different, depending on the equipment.

11) From: Frank Leaver
Jim Schulman wrote:
I wonder if the mechanism here is that the high speed air stream strips 
off and disperses some of the volatiles that would otherwise just be a 
sticky residue on the bean surface.. Even if this were true the amounts 
involved and the effect on the roast taste remain to be determined.
I've had some thoughts / questions regarding the input temperature area 
and drum vs fluid bed differences:
It's been my impression that drum roasters operate at slightly lower 
temperatures, say 470°, than small home fluid bed roasters which operate 
in the region of 520°.  While for a typical dark roast we might aim for 
a finishing temperature of say 450°+,  this is a mass average - it is 
possible that some beans or just a portion of the bean surface, just by 
random process may rise to the maximum temperature of the roaster, at 
least briefly. 
- a small portion of a bean that approached or attained 520° might not 
appear visibly charred but could it impart negative taste traits that 
would not be there if it peaked at 470° in a drum process?
- could the proportion of beans (or bean surface) that rise to the 
roaster peak vary by process or batch size? (peak to average ratio) ?  
No conclusions here.
- there is also perhaps a zigzag temperature effect at play. While we 
are generally cautioned against letting a roast cool in mid roast it 
seems to me that each bean zigzags in temperature as the roast 
progresses, with possibly detrimental effects on roast chemistry 
development for processes with more 'zag'. In a fluid bed process the 
bean enters the fluid bed, absorbs heat (temperature rise), leaves the 
fluid bed (losing heat and dropping temperature) and enters the fluid 
bed again etc. The process repeats fast enough that the time out of the 
fluid bed allows for less heat loss than is gained in the next 
insertion. The heat gain per insertion is a function of the 
bean-to-heater temperature delta and the heat transfer efficiency.  In a 
drum, heat transfer is primarily by convection (brief and somewhat 
peripheral contact of the irregular bean surface with the drum surface, 
as well as radiation and usually only a minor amount of convection). In 
both systems when a bean loses contact with the heat source, it is 
insulated from heat loss by the neighbouring beans to a similar degree, 
(although there could be a bit more heat loss for beans in contact with 
the exterior of an FR roast chamber for example). So, at the bean level, 
you might characterize roast progression  in a fluid bed as 5 steps 
forward 3 steps back while a drum might be 2 steps forward 1 step back.  
The fluid bed advances twice as fast (less time for subtle chemistry 
development) and has twice the amplitude of zag (more interruptions / 
reversals of temperature dependent chemical bond development).
All this is just hypothesis on my part of course ... none of this may be 
the case ... consider it a thought experiment. (Hey, Edison was happy 
when 1 experiment out of 100 worked ... even our failures instruct).
Comments still appreciated on this ... much to learn.
The way to test the zigzag theory is simply to develop some tiny 
bean-inserted thermocouple probes with radio transmitters ... and 
preferably with onboard GPS location modules as well.  :-)
Frank Leaver

12) From: Ken Mary
In a true fluid bed roaster like the Melitta Aromaroast, the entire bean
mass acts like a fluid supported by the rising air with little contact
between beans and metal. There is no significant zone of falling temps in a
true fluid bed. All other non-drum roasters can be considered to share both
drum and fluid bed characteristics and may have significant zones of falling
temp and brief times of contact between bean and metal that is much hotter
than the bean.

13) From: John Abbott
Ken Mary wrote:
Ken - all the fluid bed roasters I've seen (heard) have the beans 
ricocheting off the walls the entire roast. Not only do they make 
contact with the walls but do so with a fair amount of force. I've never 
seen a Melitta in action so I can't doubt your statement.  How does its 
chamber differ from lets say a Fresh Roast?   Or is the roasting chamber 
glass - and the metal (as in the FR) only in the base of the chamber - 
in which case I guess I understand.

14) From: Ben Treichel
Frank Leaver wrote:
Unmodified, your right. But with my FR, on variac I measured the air in 
temp, vs. bean mass temp. If I was pushing hard, I could run 50 to 60  F 
higher; a slow ramp, maybe as low as 10 F. The only time I saw temps 
around, or just above 500, is when I pushed the finish on a 450 F roast.
FYI, search the archives for the original posts if you want.

15) From: Jim Schulman
On 12 Nov 2003 at 12:27, Ken Mary wrote:
I have a sort of mental model that an airroaster is like a cross-
flow heat exchanger, with the air being at its hottest at the 
bottom of the chamber, and cooling to bean mass temperature 
towards the top. Individual beans themselves move around within 
the bean mass, spending time at all "altitudes."
Since the air at the coolest part of the column near the top is 
at the bean mass temeprature, I doubt the beans ever fall in 
temperature if the supply air is steady or ramped up, they just 
heat more quickly at the bottom, less quickly at the top. 
Radiation and conduction from hot surfaces may also play a role. 
But again, these surfaces will never be cooler than the bean 
mass, rather they will tend towards that temperature.
The problem with poppers and home air-roasters, as opposed to 
commercial ones, is that the path from heating coil to bean is 
very short. When I played with supply air measurements in the 
FR, I got changes from 750F to 510F in the space of about a 1/2 
inch. Andy Schechter sent readings from his PI, he measures the 
supply just before it gets into the chamber and reads around 
720F at the end of the roast. Obviously, there's usually a drop 
of 200F or so before it gets into the chamber, otherwise the 
beans would be incinerated.
This proximity to very hot air means that transient blockages to 
the airflow will create a dramatic rises in temperature at the 
bottom of the chamber. Probably one simple reason why quality 
improves with ramped roasts is that such blockages are more 
likely to occur early, when the beans are heavy, so that 
starting at a low temperature will prevent any damage. I've also 
gotten into the habit of keeping the beans moving vigorously, 
even early in the roast.

16) From: Frank Leaver
Ken Mary wrote:
You raise a good point which I missed. I wasn't considering the issue of 
bean contact with the metal components of the air roaster and, as you 
say, in most cases this contact would be minimal since the air jet 
deters bean to metal contact.
Rather I was thinking more in line with the Sivetz roaster design and the cooling effect on the beams periodically being out of the main jet stream. I was under the impression that in a Sivetz roaster the air stream / roast chamber is canted so that the beans that are drawn into the air stream are projected at an angle to the bean column mass. As beans are sucked in at the base of the air jet the beans atop descend to replace them ... the beans that have been pulled into the air stream are jettisoned to the top and back of the bean column.   
In the FR this effect is strikingly evident if the roast chamber is tilted even 5°. Beans rise up one side of the chamber and descend on the other just like they were on a high speed escalator.   In this case I would think that the beans while descending are out of the jet stream and presumably would be cooling.  
However it is true that if the FR is left totally flat the bean agitation is more random and no repetitive swirl pattern develops - rather the beans tend to bounce as a mass on the air cushion from the roaster jet. This may still lead to zigzag temp swings for the beam since those beans on the top of the pile will get less of the heat blast from the fluid bed compared to those at the bottom. The beans do move about so it evens out over time ... I was wondering though whether there is an appreciable cooling effect on some of the beans ... Maybe it's not a factor - if the effect exists it may be very small and of short duration.   
Frank Leaver

17) From: Ken Mary
The Melitta's chamber bottom is a full diameter screen of stainless mesh.
The hot air enters from the bottom, the side walls (non-stick coated
aluminum) are not heated except by the bean environment. Looking through the
mesh, you can see the mica support, there is no direct radiation from the
heating element. Most poppers have a slotted aluminum can as the roast
chamber which is heated from the side and bottom. I am not familiar with any
other roasters, having seen only pictures.

HomeRoast Digest