HomeRoast Digest


Topic: New project I'm toying with (6 msgs / 183 lines)
1) From: Obrien, Haskell W.
This one will take me a while to get done, but...
I'm once again revisiting my idea for an electronic roast profile
controller.  I'm personally very bad at sitting still and watching
coffee roast, and sometimes the 2.5year old factor strikes mid roast as
well.
A microcontroller, k-type thermocouple, relay for the heater, and maybe
a mosfet speed controller for the fan hooked up to my popper and one of
my computers (probably my Panasonic toughbook) should do the trick.
The biggest if is on the fan speed controller. I realized that there's a
simple diode bridge on the back of the fan, so I can actually use a DC
based speed controller instead of running AC power to it.
I put in a sample order for the thermocouple DAC that I need.  Once
those show up, I'll probably try to start prototyping the thing.
Assuming we've got the house unpacked by then...

2) From: Rick Carragher
Haskell,
I'm working on something similar, except I'm using a PC as the
microcontroller.  I have a type K thermocouple that will go into a
AD595 chip, which is essentially a signal conditioner that takes care
of thermocouple CJC (Cold-junction compensation) and amplifies the
signal.  Then, a labjack does the A/D conversion, that then has a USB
connection to the computer.
I was then planning on using a relay for heater, but I hadn't
considered a fan controller - interesting!
I would then write the PID code in java, or find a library somewhere. 
What I like about this idea, as opposed to a microcontroller (which I
did consider) is that it is much easier to change the profile.  The
downside is that there's more hardware (like a PC!).
I've acquired all the necessary hardware, excepting the relay, and I've
just started writing the code.  It'll be at least another couple months
before anything happens, but it's been fun so far.
Rick
--- "Obrien, Haskell W."  wrote:
<Snip>
Yahoo! Mail
Stay connected, organized, and protected. Take the tour:http://tour.mail.yahoo.com/mailtour.html

3) From: miKe mcKoffee
Whatever the mind can conceive and believe, it can achieve. I haven't gone 
that route yet know other's have already take a $1 thrift shop popper and 
turned it into a computer controlled coffee roaster. Here's Tom Gramillas...http://home.columbus.rr.com/thegramilas/coffee/roaster.htmlI recall someone else went PID control route but forget who...
Kona Konnaisseur miKe mcKoffee
3rd Annual Pacific Northwest Sweet Maria's Home Roasters Gathering infohttp://mdmint.home.comcast.net/pnwgIII.htmFrom: "Obrien, Haskell W." 
Sent: Wednesday, May 18, 2005 8:21 AM
This one will take me a while to get done, but...
I'm once again revisiting my idea for an electronic roast profile
controller.  I'm personally very bad at sitting still and watching
coffee roast, and sometimes the 2.5year old factor strikes mid roast as
well.
A microcontroller, k-type thermocouple, relay for the heater, and maybe
a mosfet speed controller for the fan hooked up to my popper and one of
my computers (probably my Panasonic toughbook) should do the trick.
The biggest if is on the fan speed controller. I realized that there's a
simple diode bridge on the back of the fan, so I can actually use a DC
based speed controller instead of running AC power to it.
I put in a sample order for the thermocouple DAC that I need.  Once
those show up, I'll probably try to start prototyping the thing.
Assuming we've got the house unpacked by then...

4) From: J.W.Bullfrog
miKe mcKoffee wrote:
<Snip>
That was me.http://www.homeroaster.com/geekmod.html<Snip>

5) From: Rich Adams

6) From: Obrien, Haskell W.
In my case, the PIC would provide a serial interface to the PC.  The PIC
would only really serve as an I/O device, rather than the actual
controller.
The brains would exist on the PC. I really have no interest in
developing a fully embedded system.  Especially since I'd like to store
and export popper profiles.
I'd considered a PIC solution, but that's a bit pricy, IMO.
The pic gives 2 way communication via TTL(Serial with a max232)
The first step for me will likely be just be the data collection
portion.
As I have time, I'll play with getting the PC to control the other bits.
My target cost is hopefully around $50 for the deal.  Cheaper using
samples.
Will


HomeRoast Digest