[Homeroast] First time post

Allon Stern allon at radioactive.org
Fri Jan 28 05:05:09 CST 2011

On Jan 27, 2011, at 11:24 AM, silas coelho wrote:
> where did you found a PID for IRoast?

Well, I'd been thinking for a while about building a PID controller for roasting; had an Arduino, found some PID algorithms, had a small pile of solid state relays and the like, but never really found the time to finish up the project.

Then I found an auction for some PID controllers and scored a great deal on a few Omron E5CK-T pid controllers.

This particular PID controller allows one to set up four different programs of ramp & soak (I leave off the soaks :) of up to 8 steps each (the manual says 16 steps, but I've only figured out how to do 8)

To understand how the PID worked (without destroying my roaster) I first experimented using a light bulb - I put a thermocouple against the light bulb, and ran the light bulb through a SSR controlled by the PID. I figured out how to run it to 100 degrees, ramp to 120 degrees, then back down to 100 degrees, how to autotune the PID, and how the programs worked. I was ready.

I  took a screwdriver to my perfectly working iRoast2. (It worked as designed, but not as I liked it - I hardly used it anymore because I was tired of having to work really hard to get any sort of control out of it, having moved mostly to heat gun roasting.) I removed all the control electronics, ran the fan wires to one power cord and both heating coils to a second power cord. I built a separate box to house the PID controller, along with an outlet box and dimmer - the dimmer controls one outlet which is used to modulate the fan speed. The other outlet is controlled by the solid state relay, which is mounted to the bottom of the (metal) outlet box for heat dissipation;  A big barrier strip handles all the 120VAC wiring inside my (plastic) controller box.

I actually cheated a little bit - the controller box I salvaged from another PID unit that I'd bought at a hamfest a couple of years ago; the PID unit that was in it was NOT suitable for roasting as it didn't do programs (though it would be fine for a single or dual boiler espresso machine; alas, I have an HX). Anyway, this box happened to have a thermocouple socket on the front, so I used that.

I run the thermocouple for the PID up through the holes in the bottom of the iRoast chamber; this means it gets a little squished by against the seals by the chamber mounting to the base. This could be a problem since I'd have to realign it every time I remounted the chamber to the base, but I never remove it - I never bothered putting the bottom half of the base back together and it's really lightweight, so at end of roast, I just remove the chamber top, and dump the whole unit over to dump the beans into the tray from the scale I use to weigh the greens.

When PID roasting, all the work that goes into roasting must be done up front - you have to think before you roast. Not just how dark do I want to roast this, but how fast do I want to get through various stages?

I have found great success using a basic profile from miKe as a starting point, and have adapted it's basic framework for other roast styles; I can see that if I had 100lbs of a coffee that I needed to roast in a large roaster, I  could use this small PIDed roaster to develop a profile for it; I don't take very good notes, however, and largely run on intuition at the moment, though I have some ideas for developing profiles that I'd like to try someday.

* caveat *

I am comfortable working with 120VAC. If you are not at all experienced with electrical wiring, either take the time to learn properly, or find someone to do the 120VAC work for you. Design for safety. This stuff can kill you.


More information about the Homeroast mailing list