[Homeroast] heat gun method

Bill S. wsmailmagnet at gmail.com
Wed Nov 24 13:30:27 CST 2010

I like your (PID) thinking, but you are miles ahead of me technically.  It
seems that you would want a temperature reading that doesn't fluctuate
rapidly, to keep the PID's output constantly moving in the right direction
(versus the PID trying to correct for a temporary temperature reading in the
wrong direction).  I have found that a probe for a contact thermometer,
buried in the roasting beans is consistent, even with manual bean
agitation.  One problem that you can have with non-contact thermometers
(depending on the roaster) is chaff accumulating on the bean surface you are
shooting.  That will dramatically throw your temperature reading off.

The S.M. digital thermometer probe that I use came with a bare bulb attached
to high-temperature wire.  I epoxied this inside an empty stainless steel
ball point pen refill, with both ends off.  When mounted with a rigid
fitting inside the roaster, this remains in the center of the bean mass, and
the SS probably provides some temperature stability giving constant


On Fri, Nov 12, 2010 at 4:03 AM, Allon Stern <allon at radioactive.org> wrote:

> On Nov 11, 2010, at 12:30 PM, Bill S. wrote:
> > At about the same point at which you are, I decided to buy a non-contact
> > infrared thermometer. Most are good to 1000F, and I have found mine to be
> > accurate in any roasting method where you can see (shoot at) the roasting
> > beans.  It has also been useful for home projects where I needed to know
> the
> > temperature of a surface.
> >
> > I addressed the air-vs-bean temperature issue by putting the probe in the
> > center of the bean mass (using SM's digital thermometer, in a homemade
> air
> > roaster, using a HG).  This way, the hot air hitting the probe has been
> > through half the beans, the beans are contacting the probe, and the
> airflow
> > over the probe is "average".  I compared readings from my non-contact
> > thermometer with the probed thermometer, and found that the tempeatures
> were
> > consistent enough for me to rely on the probe.
> I've thought about setting up a heat gun with PID, using an infrared
> thermometer as the temperature input (linear voltage or current).
> I just wonder how well the PID algorithm, as implemented in the industrial
> controllers I have, will deal with the variations induced by a hand-operated
> method, whether it will compensate adequately for them, or will be thrown
> off the rails.
> Heck, I could just get one of these
> http://www.sparkfun.com/commerce/product_info.php?products_id=9813
> and implement the PID algorithm on the microcontroller, plus an LED output
> to show process and measured temperatures.
> PID library here: http://www.arduino.cc/playground/Code/PIDLibrary
> -
> allon
> _______________________________________________
> Homeroast mailing list
> Homeroast at lists.sweetmariascoffee.com
> http://lists.sweetmariascoffee.com/mailman/listinfo/homeroast_lists.sweetmariascoffee.com
> Homeroast community pictures -upload yours!) :
> http://www.sweetmariascoffee.com/gallery/main.php?g2_itemId=7820

More information about the Homeroast mailing list