Making Your Own 57 Ounce Roaster [Updated Again]

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar

#1: Post by DJR »

Making Your Own Coffee Roaster - Thrice the capacity of a Quest for less than $253.00
March 6, 2013

See Last Post for any updates

If you have a a deck or patio that is suitable for a barbecue you can make a coffee roaster that will handle 57 ounces (all measurements here are dry measure, not weight) of green coffee which will yield about twice that volume of roasted coffee

Here is a video that gives you and overview of operation.
Here is another video that shows the details of construction.

And here is a closeup of my "Roasting Log". I'm a terrible note taker, so instead, I use removable labels that I put on the jars of beans. When the jar is empty, I take off the label and stick it to the wall for future reference.

You will need some basic skills in wood working and metal working or you will need to have a friend help or pay to have a few parts made. Or cobble things together from hardware store pipe.

The quality of the roast is very high - no tipping, roasting defects etc. The temperature is very controllable. Due to the design, there is very little chaff.

If you use the knife sharpener as a power supply you will have a good tool sharpener as well as a roaster, and you can do a bit of multi-tasking. You could get a motor instead, but the sharpener is handy to have and probably not any cheaper than buying a naked motor/reduction gear.

I chose a milk can over making a drum out of perforated metal for a couple reasons. The biggest is that I didn't want a shaft running down the center of the drum which is typical of perforated drums - the stainless steel milk can bottom is stout enough to be bolted to a flange and doesn't sag or deform. A center shaft would have meant that I couldn't use a tryer (it would bump into the shaft) and would have added weight in the wrong place for the drum assembly.

I suggest that you make the hole in the lid smaller than the one I made: I do get about a dozen beans or so jumping out per roast. Nice holes in sheet metal can be cut using electrician's punches. You could also use a small hole saw.

I used an electrician's punch to punch a couple holes in the sides. Frankly I couldn't tell the difference in the results. It does tend to burn off most of the chaff, but that isn't to be important because my cooler also sucks out all the chaff.

I originally had one vane. Then I removed it and added three smaller ones to try to reduce the noise of the beans rattling around. I didn't notice any difference sound or roast. I originally used a five gallon propane tank. I then hooked up to my 600 gallon home propane tank. I have to say, though, that a five gallon tank will last you many roasts.

Since everyone will have their own components, I will describe the principles here, not make measured drawings. Especially since none of these things is critical.

1. You will need to make a connection between the knife sharpener and the drum. A 16 inch stainless rod ½ diameter is what I use. (You could use a piece of pipe, brass, for example). I can pick up the drum by the rod and it is hardly warm to the touch. See close-ups for the two connections - one permanent which is to the drum and the other removable so you can dump the beans. I use a spider connection for the removable side and I have wired the rubber spider in place on the sharpener side so it doesn't get lost. I just drilled a hole through and ran some thin stainless wire around it.

2. Once you bolt the can to its flange and attach the shaft, you will need to make a wood bearing to hold it. This will take some measurements and maybe some shimming. I intended to use a metal bearing and just use the wood as a prototype. However, I really see no reason for changing to metal, especially since the wood is so easy to release and shows no signs of wear, some 200+ roasts later. I do spray the wood near the drum with a water bottle to lubricate it and keep it cool, but it is just one quick spray-and I've forgotten to do it and nothing happened

3. I have the roaster on a board which is tilted so the beans tend to move to the back by gravity and are forced to the front by the angle of the vane(s). You can figure our this vane angle yourself, or forget it because I don't think it matters.

4. The lid of the roaster has a ring bolt in it. That is so you can remove the lid using the tryer as a tool. That way you won't burn yourself.

5. If you don't want to machine stainless rod and flange you could use standard brass plumbing fittings, flange and pipe available at a hardware store, or even regular steel pipe, though that will rust. If you went this route, the whole thing could be made in less than a day, assuming you can make your wood bearing. A bandsaw or jig saw is handy, but actually all you really need is a handsaw and a drill the diameter of the pipe. Or forget drilling the bearing and just have it ride on the flat piece of wood with a couple screws on either side of the shaft to keep it from sliding off.

6. I mounted the burner to a piece of pipe (actually an old oxygen cylinder I cut). You could instead use a piece of steel stove pipe on top of the burner. That would work fine. You just need something to direct the heat to the drum and keep it away from the wood bearing. You could make something out of aluminum flashing and pop rivets. It isn't critical except that you want the heat directed at the drum. I've never even almost had a fire, but I keep a fire extinguisher handy just in case and I don't leave it unattended.
7. I use an infrared thermometer gun to watch the "ramp" of the roast. It doesn't tell me the internal temperature of a bean, but I don't care. I'm looking for movement, not absolute temperature readings. Correlated with sound and smell, I can tell where I am and if it is where I want to be.

This roaster isn't for everyone, that's for sure. It doesn't lend itself to automated profiles, though actually it wouldn't be very hard to control the gas via a computer. I could set this up, but have no desire to do so at this time.

It isn't very nice looking. Ok, it is plain ugly. It requires no maintenance. It has no need for a chaff collector. It has to be used outdoors or indoors only with a very substantial hood. There is no notion of controlling or adding air to the drum (whatever that does), but the heat at the drum is almost instantly controlled up or down by adjusting the burner flame. I could for example do a full roast in 4 minutes if I wanted to (I don't). You can sharpen knives - razor sharp during the 3-4 minute drying phase. Or up to first crack if you have the nerve.

I suspect that a big, powerful barbeque might do OK as a roaster. But I do see some real issues. Barbecues are basically ovens and designed to provide even heat for a large cooking surface. My design concentrates the heat exactly where it is needed. Therefore there will be less gas use and more importantly, one can almost instantly change the temperature of the drum by adjusting the burner. If I see the roast is getting a bit hot a bit fast, I can slow it down. I can turn the heat very low after 1st crack and coast into 2d. You simply won't be able to make these quick adjustments using a barbecue. I also don't like being that far from the beans. I need to hear, see and smell the roast at all times, which a barbecue won't allow you to do. Measuring the outside and inside drum temperature will also be hard to do with a barbecue and using a tryer will be almost impossible and would require opening the barbecue which will in turn let out all the hot air. Depending on how long one opened the lid, it will introduce a variable that simply isn't there in this roaster.

I usually roast twice a week. Each roast takes about 15 - 18 minutes. No cleanup, no setup, just a very simple device that gives you a high degree of control over the roasting process.

Integral to my roasting regimen is using a shop dust collector as a cooler. You can get one new for less than $90. It's basically a super powerful fan and is orders of magnitude faster and quieter (though still noisy, but the noise lasts 15 seconds, not 3 minutes like a shop vac. The use of the dust collector allows me to cool a drum's worth of beans in less than 15 seconds to ambient temperature. I suspect that an electric leaf blower would also work almost as well. What this means is that when I pull a roast, it's over. This means I don't have to anticipate and build into my roast the cooling time, which if you reach 2d crack and don't use the dust collector can really extend the actual roast time by another 2-3 minutes, especially for 32 ounces of 410 F beans. This means you can listen for the first signs of 2d crack and pull the roast at a nice C+ or if you hear 2d crack going in earnest you can stop it before you pull a Peet. Because of the cooler, I have never had a ruined roast!

5 Liter Stainless steel milk can. ... 3637373035

Price: 34.96

Slow Speed Knife Sharpener ... ner/T10097

Price: 89.95

Propane Burner ... 185&sr=8-1

Price $55.00

Stainless steel rod
Stainless steel tube for tryer
Perforated stainless sheet or stainless sheet from a baking pan or whatever.
Some nuts and bolts

Some wood

A table


I had never tried roasting more than one 32 ounce yogurt container of green beans at a time. Today I doubled the dose (actually it was 57 ounces dry measure) to see if there were any problems. The roast performed the same as half the number of beans. I had to increase the heat along the way to keep the same profile, but there was more than enough capacity. There was no burning or tipping at all. The roast was very chaffy, but as you can see, the result is no chaff at all due to the dust collector cooler.

The amount of chaff would have probably clogged a roaster that uses a chaff collector.

**************************UPDATE JUNE 4, 2011**************************

I've now done ten roasts using the higher volume (approx 57+ ounces). On windless days it works fine. If it is windy I have to max out the heat output and my roast time still goes past 20 minutes by up to five minutes. No off flavors or charring or ashy taste. But it takes too long for my attention span.

I bought a piece of 12" rusty pipe for $3.00 and sawed it in half and cut a slice out of one of the pieces.

For testing they are not connected; just resting on each other. I'll make the top hinged...or something. I'm not sure I'll paint it...I have not had much luck with so-called heat proof paints. I don't want them out gassing or peeling into my coffee.

Here is a shot showing the entire system: knife sharpener powering the drum with the temporary shroud. Below it, the gray thing is my dust collector roast cooler.

Here is a roast. I have to say that the lighting is misleading. The beans are uniform and not shiny. I may have to put up a white scrim to photograph them. All pictures are taken with an iPhone.

Here is youtube video:
The benefits of the hood are:
1. Allows my burner to easily handle the larger load with energy potential to spare. This is important if I decide to computer control it later.
2. It is much quieter. The rattling sound of the hoodless version was annoying. This is pretty quiet.
3. Wind doesn't affect it much.
4. Saves propane.

At some point, maybe I'll work on the cosmetics, once I'm completely done tuning it. Or maybe I won't.

User avatar

#2: Post by CoffeeRoastersClub »

I am trying to get the a picture of what the whole thing looks like when finished and in operation. Do you have one you can post?

"I'll quit coffee. It won't be easy drinking my Bailey's straight, but I'll get used to it." ~TV show Will & Grace

User avatar
DJR (original poster)

#3: Post by DJR (original poster) »


I just updated the post with a video (previously shown in another thread). I still need to make another couple to show in more detail the linkage and emptying the beans.


User avatar

#4: Post by CoffeeRoastersClub »

Thanks Dan. I checked it out. How well does it do with timing to 1st and 2nd crack?

"I'll quit coffee. It won't be easy drinking my Bailey's straight, but I'll get used to it." ~TV show Will & Grace

User avatar
DJR (original poster)

#5: Post by DJR (original poster) »


You do some nice metal work.

The great thing about this roaster is that as funky as it looks, it operates like a sports car.

Increase the heat, the heat instantly increases. Decrease the heat and the heat almost instantly decreases. That means that I don't have to do "mental" profiles to take into account lag time.

No other roaster (that I know of, commercial, hobby or other) other than the heat gun method which is also very good, has this instantaneous response. Even sample roasters are probably not as quick to respond. Actually I saw a similar roaster in Japan last trip which operates on a stove top and otherwise is very similar and will roast a couple pounds easily. The drum was more like the ones you make.

Here is a video of the Kamakura roastery. I may do a little article on his business -- very nice man, huge selection of green coffee and you pick your coffee and he roasts it for you. Had a little chart showing the various roasts and you pick.
So, I can decide when to dry, when to ramp to first crack, how long to wait until 2d crack etc. Sometimes if I feel first crack is a bit too vigorous, I'll actually cut the gas to almost nothing to delay 2d crack coming on too soon, letting it coast. Or even turn it off.

My times lately have been increasing. I tend to spend about 3-5 minutes below 212 f. Then I increase the heat and take 3-5 minutes to get to about 310 - 340. Soon after that first crack will start. So, say 10-14 minutes to first crack, according to what I feel like. Then I think about pulling it and usually do so about 3-4 minutes after first crack, making a total roast time of always less than 20 minutes and always more than about 15 minutes. I use a darkroom timer to keep rough track of where I am. I set it to 20 minutes and am happy to be done, including cooling and putting it in jars in exactly that time.

Len, if you were to go this route for some of your offerings, you'd have a better product than a barbecue can deliver, in my opinion, due to the ability to quickly control the heat up or down and see and hear the results accordingly.

But a commercial product would have some challenges, not the least of which is delivering a semi-open flame to the hands of people you don't know. The construction is easy, materials cheap, the Tormek is pricey, but there are three other similar sharpeners that are quite cheap. I just happened to have a Tormek gathering dust and was thinking of selling it.

The table is a metro-style shelf that I cut some holes in by cutting the wires. You would need to have some sort of holder for the gas cooker part. The stand it comes with doesn't focus the heat, it diffuses it. The linkages etc. are simple. If you offer one, please name it the "Baba Lee Roaster".....


User avatar

#6: Post by CoffeeRoastersClub »

Hello Dan. I checked out the video; neat little gear drives the guy uses to spin the roasters. The 32 oz roaster like you have there is a great size for personal use. When roasting for myself that is usually the amount I roast.

Also, thank you for the kind comments on the work I do. :)

"I'll quit coffee. It won't be easy drinking my Bailey's straight, but I'll get used to it." ~TV show Will & Grace

User avatar
DJR (original poster)

#7: Post by DJR (original poster) »

I've made some updates to the original post showing a new hood etc.


User avatar
DJR (original poster)

#8: Post by DJR (original poster) »

I decided to replace the wood framework with a stainless steel one. The wood one worked fine and showed no signs of wearing out. However, I'm trying to learn how to weld and this seemed like an easy project, which it was. I used a high temperature (unnecessary) pillow block (about $15) bearing which is held loosely by two stainless pins. Cost of materials for this upgrade was about $35.00.

Here is a video:

User avatar
DJR (original poster)

#9: Post by DJR (original poster) »

Dumping the Roast

I've been asked how I do it, and since it takes two hands, I had to wait until someone could make a video. It isn't very elegant, but has worked fine for a few hundred roasts (so far I haven't dumped them on the ground). Anyhow, elegant or not, it is fast. The shaft doesn't get very hot.

User avatar

#10: Post by CoffeeRoastersClub »

DJR wrote:Dumping the Roast

I've been asked how I do it, and since it takes two hands, I had to wait until someone could make a video. It isn't very elegant, but has worked fine for a few hundred roasts (so far I haven't dumped them on the ground). Anyhow, elegant or not, it is fast. The shaft doesn't get very hot.
I watched the video. Looked good to me! Quick and efficient.

"I'll quit coffee. It won't be easy drinking my Bailey's straight, but I'll get used to it." ~TV show Will & Grace