After much reading on this and other sites and playing with extension cords and other tricks to get better roasts out of a Poppery I I decided to go the PID route. This is where I landed and how I got there. I am able to consistently duplicate the fluid bed profiles described on this site with a first crack at 390 around 9 1/2 minutes and an end of roast about 30 seconds into second crack at 455 between 13 1/2 and 14 minutes.


Roaster running. Note that it is vented directly outside such that smoke and chaff are directed out of the basement.
Roaster components:

• Poppery I with fan and heating circuits split to allow descrete control and thermocouples to monitor input air and bean mass temperature.
• Variac voltage controller to control fan speed (0 - 120% of normal speed).
• Control box to monitor roaster and regulate input air temperature (two PIDs, Solid State Relay, etc.). The controller used provides for computer connection but I've not found that necessary.

Controller showing components used. The small 'basic' PID just monitors the bean mass temperature. It controls nothing. It was just the least expensive way I found to monitor the output of that thermocouple (~$35). Note that the Ramp/Soak controller shown is an Auber. It did not allow for sampling at a rate fast enough to prevent wide input temperature fluctuations. It was replaced with a Delta model that works very well.


Controller showing connections on rear. The thermocouple socket connections and plug for heating circuit allow the controller to be easily separated from the roaster.


Placement of bare K type thermocouple in the heating element chamber between the heating element (not shown) and the fins where air enters the roast chamber.


The heating chamber thermocouple cable is secured to the roast chamber housing to prevent dislocation or stress at the point where it is epoxied in place.


The bean mass temperature is monitored via a second K type thermocouple. This is a two inch long water tight component mounted about two inches below the top of the bakelite portion of the chamber. The component came threaded so it was easily secured with a nut.


The roaster was mounted on a small purpose built shelf on a hinged wedge. It is held in place by a pair of turnbuckles and hooks fabricated from clothes hanger wire. Tilting the roaster toward the wall during roasting facilitates bean circulation and allows for a larger quantity of beans to be roasted than if the roaster was plumb. I have found that a 6 oz. load circulates well and roasts more evenly than smaller batches. I've heard of people roasting 8 oz. with this arrangement but I've not tried. The glass chimney is from a replacement FreshRoast chamber. They are easily found and provide visibility while keeping beans from being thrown out of the roaster as the mass expands and becomes lighter. The chimney is held in place with springs and hooks from a plate hanger. Works great.


Once the roast is complete the roaster is brought to level (not obvious in the photo) and the fan speed can be increased without risk of throwing beans out the top. This allows rapid cooling. Note that the exhaust elbow has been removed and the hole to the outside covered. Someday I'll get real thumb buttons.


The hinge also allows the completed roast to be dumped by simply swinging the roaster down. This can be done with one hand and keeps the cords from being stressed or twisted to a point where they would deform.

Update: After replacing the PIDs with new ones from Delta I decided to take advantage of the serial output and add an interface to a computer to better monitor roasts. The graph displays the temperature of the two thermocouples as a function of time:


Also, I fashioned a cooler from a high velocity cabinet fan, a strainer, and a plastic detergent jug. This allows me to cool my roasts to room temperature in 30 seconds or less (depending on the temperature in the basement).

That's awesome!

Can you really roast 6 ounces with just the addition of the glass chimney and tilting the popper?

+1 on Awesome. Let's see you top this with your second post! LOL. Welcome aboard.

Where did you source those sockets for the thermocouple mini plugs?

Wow, I honestly get jealous when I see people trick out poppers to this degree. I would love to have simply modded popper to compare fluid bed vs. drum. Some one needs to sell a kit for something like this...hint...hint! Nice job!!!

Nice setup! And just think, if you ever get bored, you can put popcorn in there and really mess with your neighbors! Just picture popcorn rocketing out the side of your house to your amazed onlookers.

mborkow wrote:Can you really roast 6 ounces with just the addition of the glass chimney and tilting the popper?

Yes. Tilting the popper works by creating a a bean mass that is more thicker on one side and thinner on the other. The thinner side of the mass yields more easily to the pressure from the fan resulting in bottom to top circulation of the mass instead of a vortex like swirl. A chimney is required only to keep the beans in the popper. The material isn't important. I was using a soup can with both ends cut out before I moved inside and added the venting elbow which made it impossible to see the beans. The use of a bicarbonate glass cylinder solved that.

tamarian wrote:Where did you source those sockets for the thermocouple mini plugs?

Auber Instruments (http://www.auberins.com/index.php?main_ ... cts_id=119) for $5.89. I sourced most of my components from them and was really happy with their quality and service. The ramp/soak PID I got from them wasn't fast enough to control the heating element but they told me up front that they didn't think it would be. I never imagined that the NiChrome element would be so responsive.

Sincere thanks for the comments. I was a bit reluctant to post since the only thing I did that I hadn't seen before was the hinged mount.

Best,
Ed