HomeRoast Digest


Topic: New roaster (person, not appliance :-) (28 msgs / 739 lines)
1) From: Jim Karavias
Hi,
I'm Jim Karavias and I have recently begun to roast my own coffee.  I
originally started because I found out that I could get very high quality
coffee for much less than I'd pay if it was already roasted (important to a
dot-bomb victim), plus I'd know exactly when it was roasted so could be sure
of freshness.  What has really hooked me is the blend of technical knowledge
and subjective aesthetic judgment that can brought to play in this process -
in whatever balance one likes.
I've been doing some research into the possibility of making my own fluid
bed roaster (I'm sure you've all heard that before :-) and I was wondering
if anyone had come across a resource that described how to calculate the
range of air flows and pressures required to fluidize a given sample size.
I've come across process air heating equipment and blower suppliers and I'd
like to be able to ask for quotes and be fairly specific about my
requirements - plus I just like the physics.   Any pointers would be greatly
appreciated.  For those of you with a similar bent here's a URL I found
interesting:http://www.niroinc.com/html/pharma/pairflowarticle.htmlFocused on the pharmaceutical applications of fluid bed processing but
hey...
In the meantime I'm happily roasting with my Toastmaster (salton) corn
popper with quite satisfactory results. I never would have thought that with
such little effort I could produce such high quality coffee.
Regards,
Jim
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

2) From: Dan Bollinger
Jim, Welcome!  I've toyed with the idea, too.  From what I can tell, you
need the higher pressures developed by a centrifugal blower like those used
in the smallest of shop-vac. It would take a very large diameter squirrel
cage blower to raise beans. Dan
<Snip>
a
<Snip>
sure
<Snip>
knowledge
<Snip>
process -
<Snip>
I'd
<Snip>
greatly
<Snip>
with
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

3) From: Jim Karavias
Thanks Dan, thats good to know. I've been guessing the static pressure for
the fan to overcome would be at a minimum .5 lbs/ 12 sq inches (4 inch duct)
which is just about .05 psi.  Which doesn't seem like much pressure so I'm
inclined to think I'm not going about this correctly.  Ideally I'd just set
up the blower, heater and duct work and tweak it.
How far did you get with your idea?
Jim

4) From: Ed Needham
Your calcs may be right if the mass to be lifted was solid, and lifted all at
the same time.  Since the beans move around, the pressure is allowed to
'leak' through the beans, so in practice, the air not only needs the static
pressure, but also volume and velocity.  A squirrel cage blower can do the
velocity and volume, but fails miserably on pressure.  A fluid bed roaster I
started, but have gotten sidetracked on, is built with a Fasco 365 CFM
squirrel cage blower.  I 'think' it will work, but in case it doesn't, I've
also purchased a high static pressure Ametek vacuum/blower, similar to this,http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item71546133but without the
built-in motor.
I need to buy a separate belt drive motor to run it, but it 'will' do the
trick if the Fasco can't.  I'm not a fluid dynamics engineer, but I can
usually SWAG with the best of 'em.
Ed Needhamhttp://www.homeroaster.comed
****************************************
**********************************************

5) From: Scott Jensen
The Syd and Jerry's roaster which holds 1.36lbs of greens uses about a 6"
squirrel cage fan.  It is enough power if everything is perfect.  Have to
keep the vanes, and all components cleaned though.  Not much tolerance.
Scott

6) From: Dan Bollinger
<Snip>
Scott, That's good to know.  The larger the diameter of the squirrel cage
the greater the pressure. The down side is that the increased airflow is
cooling down the heaters.  I've set-aside the notion of a fluid-bed roaster
because of the noise factor when using a centrifugal blower.  A squirrel
cage blower is much quieter.  I'd still like to do a hot-air design, because
it is so easy to monitor and control.  I don't have anything solid yet.  I'm
keeping everything undecided and seeing what gels.  Dan
PS:  Drinking a great La Berlina 11m City roast this morning.  :)  Dan
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

7) From: AlChemist John
Sometime around 20:54 9/29/02, Jim Karavias typed:
<Snip>
I was looking into those same calculations myself.  The equation you need 
is known as the Ergun equation.  It is relatively simple, but not very 
helpful.  It was a rather empirically based equation.  In order to use it, 
you need the terminal velocity of the object you are trying to fluidize and 
to get that value, you need it's (the coffee bean's) coefficient of 
friction and shape "constant".  As MANY people (engineer co-workers) told 
me, it is a no win situation for use on a small scale.  In order to design 
and build a fluid bed, you need the "constants and coefficients".  In order 
to get the "constants and coefficients" you need to build a fluid bed and 
get the numbers experimentally...sigh
I didn't like that solution but that appears to be the way it 
is.  Basically, reverse engineer an existing system for basic 
"guessimates", hook up a blower and see if it works.  If the beans hit the 
roof, get a smaller blower (or roast more), if they don't move, get a 
larger blower.  Good luck.
I will check out the URL.  Thanks.  Welcome aboard.
--
John Nanci 
AlChemist at large
Roasting and Blending by Gestalt
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

8) From: EskWIRED
<Snip>
One thing I've seen people do in designing home made fluid bed roasters is
to use a cylinder as the roasting chamber.  I've heard reports of beans
being propelled out the top or the cylinder.
I read the original patent for the Popcorn Pumper, which was the first hot
air popper.  They made the point that they used a chamber which was larger
on top than on the bottom, which allowed the pressure to dissipate as the
air moved upwards.  I use a hurricane lamp chimney to extend the roasting
chamber in my WEPP, and I've noticed that the beans never get blown up
further than where the glass chamber expands.  I've also noticed that big
giant commercial fans used for fluid bed processing have a flared shape,
which I presume is for the same reason.
I think that this design point is important, and would allow the use of a
much stronger fan without the shotgun effect forcing you to turn down the
fan.  Somebody recently posted a URL to a technical paper on fluid bed
drying and coating (I think it was whoever started this thread).  I read it
yesterday, and it is VERY informative.  I'd recommend reading it and
re-reading it.  It is athttp://www.niroinc.com/html/pharma/pairflowarticle.html It is noteworthy
that all the diagrams in the paper show flared chambers.  Look at the flared
chamber at the bottom of the fluid bed dryer on this page:http://ace.phys.h.kyoto-u.ac.jp/~ichiki/test.htmlOne interesting aspect of the paper was the different types of air inlets">http://www.umangpharmatech.com/p_fb_bot.htmThis page has some ultra-cool animations of fluid beds:http://ace.phys.h.kyoto-u.ac.jp/~ichiki/test.htmlOne interesting aspect of the paper was the different types of air inlets
used in commercial processes, and how they affect the airflow.  Hearthware
uses a vertical airstream in the middle, while poppers use a rotating stream
on the edges. We should think about the pros and cons of these two
approaches and figure out how to best use them.
Another thing I found interesting was a short mention of the ratio between
the diameter of the chamber and the height of the fluidized bed.  Again,
Hearthware uses a flatter bed, while poppers use a tall skinny bed.  I would
think that would make a huge difference in whether beans get blown out the
top or not.
I am not an engineer (IANAE), so I could be totally wrong about everything
:)
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

9) From: Jim Karavias
Hi John,
So, Roasting and Blending by Gestalt could be augmented with "Building".
I'm afraid that's really what it's going to come down to.  I don't have
access to a fluid bed roaster to reverse engineer but there are some
close-ups of Sivetz designs that I could use as a guide. (Not really
interested in competing with you Mr Sivetz - purely a fun endeavor :-)
There is a process air heating company that lists fluid bed roasters as
applications of its blowers and air heaters...
check out:http://www.malcom.com/pdf/gb_lg69.pdf (they sell Leister process
heating equipment)
and look at page 24, right column middle picture. Not a Sivetz machine but
I'm sure I've seen it before in my browsing. It specifies the heater and
blower used for a fluid bed roaster.  Its just a much larger scale than I'm
interested in.  I got a message a sales person from Malcom industries and
will follow up with them regarding which of their equipment they beleive
could fit the bill.
My 'research' certainly has turned up information that explains why these
machines are so expensive at the small commercial level.  I had no idea that
such a blower and heater combination, without the labor and supporting
machinery/control circuitry, could cost in the $1000 range if purchased in
single quantities.
I suppose I'll need to start a coffee roasting coop and charge membership
dues to pay for a nice small commercial machine...  who's in :-) ?
Jim

10) From: Jim Karavias
That's a great article isn't it.
I know this moves pretty for OT from coffee roasting...but I have a friend
at Lawrence Berkeley labs that uses fluidized beds for geological studies
since fluidization can occur as a result of vibrational forces as well as
air flow.  Sand and fill seem to behave as fluidized beds during an
earthquake.  I'll have to time my big roast with the next 'Big One'.
Jim
========================================================================
Somebody recently posted a URL to a technical paper on fluid bed
drying and coating (I think it was whoever started this thread).  I read it
yesterday, and it is VERY informative.  I'd recommend reading it and
re-reading it.  It is athttp://www.niroinc.com/html/pharma/pairflowarticle.html It is noteworthy
that all the diagrams in the paper show flared chambers.  Look at the flared
chamber at the bottom of the fluid bed dryer on this page:http://ace.phys.h.kyoto-u.ac.jp/~ichiki/test.htmlOne interesting aspect of the paper was the different types of air inlets">http://www.umangpharmatech.com/p_fb_bot.htmThis page has some ultra-cool animations of fluid beds:http://ace.phys.h.kyoto-u.ac.jp/~ichiki/test.htmlOne interesting aspect of the paper was the different types of air inlets
used in commercial processes, and how they affect the airflow.  Hearthware
uses a vertical airstream in the middle, while poppers use a rotating stream
on the edges. We should think about the pros and cons of these two
approaches and figure out how to best use them.
Another thing I found interesting was a short mention of the ratio between
the diameter of the chamber and the height of the fluidized bed.  Again,
Hearthware uses a flatter bed, while poppers use a tall skinny bed.  I would
think that would make a huge difference in whether beans get blown out the
top or not.
I am not an engineer (IANAE), so I could be totally wrong about everything
:)
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast">http://lists.sweetmarias.com/mailman/listinfo/homeroasthomeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

11) From: Dan Bollinger
Jim,  I did some research on this awhile back and found the Leister
machines, too.  They are expensive.  Another option are the smaller and
cheaper hot-air guns byhttp://www.masterappliance.com/ Sivetz uses one in
his sample roaster. I have a Master Appliance hot-air blower (very quiet).
Not enough for fluidizing, but would make a good heat source for roasting.
Got it for $39 on ebay!  ;)
Also, Sivetz' patent is a good source of information.http://patft.uspto.gov/netacgi/nph-Parser?Sect1O2&Sect2TOFF&p=1&u=/netahtml/search-bool.html&r=1&f=G&lP&co1
D&d=pall&s1='3964175'.WKU.&OS/3
964175&RS/3964175
And, Gardfoods Roller-Roaster, twin chamber design.http://www.gardfoods.com/coffee/coffee.roaster.htm<Snip>
process
<Snip>
I'm
<Snip>
that
<Snip>
wondering
<Snip>
size.
<Snip>
I'd
<Snip>
and
<Snip>
order
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

12) From: Ben Treichel
Dan Bollinger wrote:
<Snip>
Thats what I figure I will use. However, I intend to design the 
controller first.
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

13) From: Dan Bollinger
<Snip>
What kind?  Analog?  Digital?  Programmable?  I'm thinking of going analog,
watching the roast as it goes and manually controlling the heat.  With
larger batch sizes I won't mind the 20 minutes.  It will be a sort of weekly
meditation. Dan
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

14) From: Ben Treichel
A programmable microcontroller based one. I think I will use, at least, 
two thermocouples. Since I can either put in two ice point/amps for the 
TC's or put a  8 channel multiplexer in front, I figure I will go with 
the mux. I am figuring on an relay or power FET that will handle at 
least 1500 watts, probably a 15 amp 120 volt part. But now as I think of 
it, a relay might be better. Easier to change the part to a bigger size. 
I guess size and cost will be questions.
I figure I will control air temp below the beans, and monitor the bean 
temp. However, TC placement will be an issue. Not so much from a 'real' 
temp standpoint as repeatability. Another question is should I look at 
an IR sensor for bean temp. But the cost is 20X's that of a TC, and I'm 
not sure it would improve repeatability.
I'm figuring one box with a couple line lcd display, and a keypad, or 
maybe a pc keyboard. That way, you can enter the roasting profile such 
as , 350 at 0 time, 375 at 3:30 etc. probably just ramp the temp between 
points, and to hold temp just put in 375 @ 9:00 375 @ 12:00.
Also, trip into the cooling cycle at a set bean temp. Allow 10, or 20 
roast profiles to be stored. I will undoubtedly connect to my PC since 
that the development system host, but want it to be standalone.
I'm not concentrating on the roasting chamber right now, because I 
figure I'm planing on interfacing it to current home roasters. Like my 
new FP+. Besides, I only use about 1/2 lb a week, so a 3 oz roast is 
just fine with me. However, I think the guts from multiple air guns 
could be used for a 'staged' roaster. I.E. kick one in for 3 oz, 2 for 6 
oz, and 3 for 9 oz. That does raise some interesting chamber design 
questions.
Ben
Dan Bollinger wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

15) From: Dan Bollinger
<Snip>
I think a two TC control is ideal.  You'll need analog cartridges/inputs for
the TC's right?  A relay is the most common in electric ovens.  For very
fine control I've found that a 20-60 second scan cycle is a good balance
between temperature control and relay longevity.  Make sure you get a HD
relay, look up the 'resistive amps'. I'd get the next larger relay size, but
with your weekly roasts it might not matter.
<Snip>
Go TC.  Very accurate and very dependable.  I've seen them last 10 years in
constant use.  Just work on finding the sweet spot in the roaster.  I'd use
both TC's to control heat.  Use the heater TC as a 'do not exceed' value and
use the bean TC as a 'target/process temperature.'  In other words, wire
them in series with the relay.  Or, run two ramps simultaneously.
<Snip>
OK, here's a stupid question.  When you press the start button, how do you
get the temperature to be 350° instantaniously at 0 time?  The Pro1500 has
eight programmable ramps. I suppose you could always buy their board!
<Snip>
Actually, you don't need that much more heat for the larger roasts.  It
isn't linear since most of the heat goes up the chimney.
Dan
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

16) From: John
Ben,
Love your approach.  I built an interface to my PC with limited control
on the application board. I don't have much of a lab anymore and ran
into trouble with the triac switching and finally went with relays to
control the Fresh Roast electrical system. I used a couple of LM339's
with signal conditioning for the thremal probe which I mounted in the
air flow at the base of the removable roasting chamber.  I now think
that a second probe at the top of the stack might have given me more
control.  BUT.. I can tell you a lot of things NOT to do and what
happens when you do :O) 
John - loving life in the slow lane
On Wed, 2002-10-02 at 09:16, Ben Treichel wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

17) From: Ben Treichel
Dan Bollinger wrote:
<Snip>
  No. 15 dollars in parts gets me everything in the input modules. I 
then just feed this signal to the a/d on the chip.
 From what I've seen they want about 100 for a single module.
<Snip>
Ah, if they wear out from cycling I may go with an on board FET/power 
darlington. They don't.
<Snip>
I plan on monitoring both loops, with a manual override of course.
<Snip>
I'm not, its just the first setpoint, the heater will come full on until 
its meets that point, I'm not sure that I will limit the number of 
points in the profile, but I could set the 1st point as the warm-up and 
hold point.
<Snip>
Just more fan to keep the beans aloft?
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

18) From: Ben Treichel
John wrote:
<Snip>
Okay, looked up the part.
What I'm looking at using is:http://products.analog.com/products/info.asp?product­595Monolithic Thermocouple Amplifier with Cold Junction Compensation">http://www.atmel.com/atmel/acrobat/doc1939.pdfthis is the dev board.http://products.analog.com/products/info.asp?product­595Monolithic Thermocouple Amplifier with Cold Junction Compensation
and there is a paired mux.
Not sure on the output control yet, but the control will be bang/bang.
I'm no electronic design wizard, so it will be a little touch and go.
<Snip>
That would be great! We have a saying where I work, education is 
expensive, no matter how you get it.
<Snip>
wish I was there. another 10 to minimum 58.
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

19) From: Jeffrey A. Bertoia
Ben
Have you considered TINI from Dallas?
Be pretty good to have a roaster written in JAVA with access network wide
on a browser.  Seriously with TINI and a Step+ from Systronix (it gives you
the prototype area for the TC controller and the High Current IO for your
relays you would be about done!
Seehttp://www.ibutton.com/TINI/index.html.I have been using this for a number of clients and it is definately Slick.
jeff
Ben Treichel wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

20) From: John
Jeff,
However it would still require an interface between the Tini and the
Roaster. Switching the triac/relay and getting the information on
temperature out of the box.  What the Tini would accomplish would be
making the control of the final product available to a network. 
John - wishing they would bring back the piggy-back i8047 - life was so
easy with that chip.
On Wed, 2002-10-02 at 14:50, Jeffrey A. Bertoia wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

21) From: Ben Treichel
Jeff
I did a quick look at it, and seems to be a nice architecture, but I 
haven't found 1 wire, etc, sensors that go above 100c. Also, I would 
like to keep the parts cost in the 50 - 60 dollar range. If I was doing 
this as a job, I wouldn't care. But we start talking about REAL money, 
when its mine.
If you want to point me in the right direction for parts, I would be 
interested in looking at them
Thanks,
Ben
Jeffrey A. Bertoia wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

22) From: John Wanninger
How about using Lego Mindstorms? The Lego Mindstorms Controller w/ROBOLAB
software will  operate over the internet.  The conroller has 3 PWM outputs
and 3 multi-purpose inputs.(The 10 to 12 year olds on the two First Lego
League teams I coach are becoming programming whizzes on these things!!).
Anyways this controller could drive a solid state relay or the small DC
motor on my motorized  0-180V 20A "Variac" (often available for $45 on
ebay).  Use 1 input for A/C voltage(scaled and rectified), 1 for air inlet
temp (RTD) and 1 for bean/air outlet temp (RTD). Besides the ROBOLAB, Many
languages (and operating systems) have been developed for this controller
including FORTH, JAVA and NQC (Not Quite C).
John
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

23) From: Jeffrey A. Bertoia
Ben / John
I would use the TINI along with the systronix step+ socket board...
(for a prototype).http://www.systronix.com/tini/step_data.pdfThe step+ has an on board 4 channel 16 bit adc.  The DS2450 (one wire)
you could then use the AD  part you mentioned earlier as a TC interface.
The step board also has a prototype area to mount this part and the control
stuff.
The step+ also has high current IO capable of driving relays or SS Relays.
Yes it is definately overkill but the reduction in development time would be
incredibile.
In response to John...  The TINI is quite capable of control...  In fact 
your
longing for the 8047 is amusing...  The 80C390 (the processor on TINI)
is really nothiung more than an 8051 on some serious steroids.
If anyone would like to continue this off list I have already put a lot 
of time
into it and would be glad to share my research or be part of a development.
jeff
Ben Treichel wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

24) From: Dan Bollinger
Tell me more!  What's their URL.  A forum?  Dan  (who loves low-tech)

25) From: Ben Treichel
Have one. Look here. There is also a hacking site, lost the url.http://www.oreilly.com/catalog/lmstorms/Dan Bollinger wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

26) From: Ben Treichel
yeah I'm interested, anybody else want to think about coming along for 
the ride?
Come one guys, at least give it a sniff.
Talk to you off list Jeff.
Ben
Jeffrey A. Bertoia wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

27) From: Mike Gallant
	I'm in. I'm not much of a hardware guy (willing to learn), but I 
have some embedded software development experience (and an old roaster 
or two to experiment with)
-mike
--
Mike Gallant
pischke
On Wednesday, October 2, 2002, at 08:08 PM, Ben Treichel wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

28) From: Ben Treichel
great, should be fun, and one of the reasons I'm doing this is to 
'relearn' what I knew.
BTW, I knew a Mike Gallant at Cummins.?
Mike Gallant wrote:
<Snip>
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast


HomeRoast Digest