Greetings all,

I'm attempting to find some books about building a home roaster. First and foremost, I would like to find books light on the art and heavy on the science of roasting. I have found several books about the art of roasting and fill confident enough in such an area. I need to now get down to the 'brass tacks', as they say.

I'm looking to build something similar to a Probat style roaster. I don't need anyone to attempt to discourage me from even trying such a harebrained idea; I'm surrounded by said people. It has been done before ( I used to live next to someone who built one, but he is long gone) and I be believe I can do it. I have two fabricators willing to help me. I'm up on the art of roasting, but I really need a book inclined toward the science(i.e. thermal dynamics, chemistry, etc.) I don't know if I'm looking in the wrong places (local libraries) or if they're out there just difficult to locate.

Any help would be greatly appreciated. Thank you in advance.

I don't know of a book but a number of drum roasters have been built and discussed on Homeroasters.org
Building a Roaster section

The Illy espresso book contains a good chapter on roasting in general. The drawings for the Probat L roasters are public domain, and you should find them with a Google search.

The basic Probat and Gothot design is that the drum is solid and ventilated. The beans are rotated towards the front. Air is forced over the heat source and then both through and around the drum from the rear to the front. On Probats, there is a damper that can proportion the air through or around the drum. Similar designs do not use this. There is also a control to vary the overall amount of airflow.

The air is not recirculated, and the airflow is not high enough for the system to be considered a pure convection roaster like designs patterned after the Burns Thermalo (e.g. the Loring roaster). Instead, the airflow is similar to the natural convection one gets through perforated drums when these are gas heated (i.e. roasters patterned after the Burns Jubilee design).

The threads on this forum about the Quest M3 and Mini 500 give an idea of the detailed design of some shop built Taiwanese ventilated solid drum roasters. These are well executed designs by enthusiasts, but the detail differ from Probats.

I've looked high and low for books on building home/sample roasters and if they exist, they're well hidden.
I agree with Ed on checking out Homeroasters.org. I've seen some great builds coming from them.

I've built a "Probat style" 1/2 lb capacity drum roaster and recently a 1 pounder with additional inside-the-drum quarts radiant. Here are some pics of the drum and front plate of the 1/2 pounder.



The drum is perfed but I will be modifying to solid fairly soon as I've found the solid with heat flowing from underneath and then through the back over the beans and out the upper front to produce superior results.

My hope is for you to see the basic construction of the agitation vanes within the drum. The beans rise fast enough to fall about mid-drum (80 rpm for mine) and are constantly moving toward the front of the roaster due to the vanes' helical shape. You'll see the front half of them are trimmed down in height to allow ample clearance of the bean trier and bean return plate. The angled sheet metal plate you see on the back side of the front plate (return plate) takes the place of the shaft mounted reverse helical vane assembly in larger roasters which recirculates the beans back to the rear of the drum.

The drum is supported only at the rear hub since using a three spoke spider would have interfered with the return plate.The rear hub is a gear blank with set-screw mount and has a 5/16" bore shaft coupler welded to it. I added three set screws to the opposite end of the coupling and drilled out the bore to allow centering and allignment of the drum.

I had a local sheet metal fab shop roll the pre-perfed 16 gauge drum and weld everything up after I cut out the raw parts I was capable of. They built the charge funnel from scratch.

I had them tack weld a set-screw shaft collar to the front for securing my bean temp sensor. I cut off a length of .125" dia stainless tube from an old bi-metal thermometer to allow affixing my thermocouple sensor 2" inside the drum. The location allowed the beans falling from the return plate to constantly hit the end of the thermocouple. The thermocouple tip was not within the stainless tube but was sticking out the end to allow direct contact.

For shaft bearings you can use ball bearings but I decided to use carbon composite bushings mounted in shop-fabbed brass journals which you can see in the pic showing the front plate. They can handle all the heat you could ever give them without damage but with ball bearings you need to space them out away from the front a bit to lessen direct heat transfer which causes grease separation and dripping down the front. Some folks have kept them direct to the front but had to use a high temp grease which will sometimes separate at low rpms and cause dry pockets.

The trier was really a waste of time as I never use it. I, like most of us, use temperature, smell and C1/C2 sounds for judging end of roast. It is sometimes nice to see the color transitions to yellow-yellow/brown which the trier affords.

The heating elements (1000 watts) just underneath the drum are modeled after a Probat RE-1 sample roaster which resemble bread toaster mica-board panels around 2 1/2" tall but with 1/4" dia coiled nichrome wrapped around the mica board (3/8" spacing between wraps around the mica) instead of flat nichrome ribbon. The coiled nichrome causes much better heated air upward convection than flat ribbon if you are using it without powered exhaust. If you use powered exhaust back to a chaff separator/exhaust fan combo the coiled nichrome transfers sufficient heat to the air stream to allow 400 F plus air to be entering the back of the drum and across the beans (An air path is necessary below the heating element area from the bottom of the roaster). Roasters using tubular heating elements get the job done but typically don't transfer enough heat to allow hot enough air to be contributing to the convection part of the roast and rely mainly on drum heat. The Quest M3 may be a different animal in that respect from reading Arpi and Jims posts.

Environment temperature is sensed just prior to the air entering the back of the drum (1/2" from the drum near the shaft). It will go from 300 F to 480 F during a roast. I will be modifying my charging funnel to allow an exhaust tube to connect to it just below the square cone portion and will travel back to an exhaust/chaff collector. A damper will be placed in-line to set air flow level.

Motor drive is a Merkle Korpf 120 rpm C-frame gear motor using miniature pitch timing pulleys/belts (neoprene). The pulleys were sized to reduce down to 80 rpm.

Drum size is 4 3/4" dia by 6" long out of 16 gauge mild steel. Stainless can be used and has a slower heat transfer rate which is probably not a bad thing. If I were to do it again I would make it 6" dia by 6" long for a 8 oz charge weight. Most Probats have a drum length close to its diameter.

I had a local gunsmith blue the front plate parts after I hand sanded and polished close to a mirror finish. The bluing is nice but if you have it in a real high humid environment where it can hit due-point there is a chance of rusting. Here in Colorado it isn't a problem. When I had it in Virginia it rusted slightly in a couple of areas.

Hope this helps. I can answer any questions you might have in regard to my rigs any time.

Cheers

Allen,
Beautiful work. I have a similar design I hope to commercialize.
Jim, could you please explain how most of the larger roasters control temperature? I've seen the video on the sample roaster from US Roasters that has a PID controller. But it also has an adjustable pressure regulator on the gas, as well as the PID, a damper, and a vent fan. Are they adjusting the vent fan speed, or controlling the gas valve, or just moving a damper on the vent? I'm confused how these things are working together, and am sure others are wondering this as well.

Thanks,
(Yet another) Jim

I do not know how good shop drum roaster PID controls work. Large industrial roasters use a dual loop controller where the output of the bean temperature PID changes the set point on a supply air temperature PID, which in turn regulates the heat. These are tricky, the controls version of using a compound pendulum; but they can handle long lag times more easily than single loop controllers.

As far as manual control. Roasters turn the gas up and down. They usually have multiple thermometers for information: I've seen bean temperature, drum temperature, flue temperature and exhaust temperature. They may have gas flow meters as well as pressure, for better precision.

Using the dampers appears to be a bit of a black art. Opening them fast can cool off the roast, but opening them up with added heat can speed the roast up. On my small Probat knockoff, I measure the drum/intake air temperature. There is a very small increase in roasting speed when one increases air flow, keeping its temperature the same; but I have no clue how great the effect is on other roasters with this layout.

If a pro has ever used more than one drum roaster, he or she will not to say anything about controlling a drum they don't know -- so I suppose that most of the know-how for these beasts is very local indeed

Thank you Jim,
Roasting is like poker. Quick to learn, but never quite mastered. Do you have any references you might recommend for control systems that would be applicable to these roasters?

Jim

The best introductory chapter on roasting theory is in Illy. It covers the basic roasting devices and profiles, heat transfer, water transport and some chemistry. It also contains a complete list of references. Unfortunately, there won't be much on controls that is explicit, since this is mostly proprietary. This is all academic literature, so you need access to jstor, etc, via a good library

Controls are controls; the same devices are used for coffee roasting or making pots; it's a question of understanding the different types of controller and how to tune them. There are lots of textbooks, but they require college level algebra.

You don't need any of the math to build a PID control for coffee roasting. If your heat is gas, you need a 4 to 20 ma actuated valve, if your heat is electric, the proper size SSR. Since the quality of the roast depends on both bean and environmental temperature; and since environmental temperatures will oscillate if you run a PID controller on bean temperature, the most reliable control is to have the PID controller steer the environmental temperature in a ramp that produces roughly the right bean profile, and have a second on/off temperature switch end the roast when the desired bean temperature is reached.

If the last paragraph doesn't make sense to you, spend a day googling PID project; since it's impossible to explain more in a single forum topic.

Thanks Jim,
The information makes sense. I will do some searching and see what I can find.

Jim

Wow Allen, great pictures and machine.

I'm scoping building one of these to replace my popcorn popper. Luckily I have lots of metal fab and machining equipment, PID controls, nichrome, etc so it should be all doable in-house.

Mind some questions?

How big did you make your drum for the 1lb unit?

It seems like the main difference between the Quest and yours is the build quality (yours is built heavily and with much more care), the heating element design and airflow.

Having never seen a drum roaster in person it's also difficult to visualize how the drum seals up against the front and rear plate. And I'm having a hard time understanding how air flows through the drum if it's sealed off at the front.