Prototype Turbo Oven Coffee Drum Roaster

IMG_20180204_135918Inspired by my friend Dominick’s attempt at a Turbo Oven-powered Drum Roaster I thought I’d build my own.  While I appreciate the simplicity of my Stir Crazy base I’ve never been happy with the consistency of roasts; by its very nature it just can’t stir the beans around to evenly roast them.

In his version, Dominick talked about a long roast time of about 20 minutes.  I figured that was due to heat loss through the sides of the enamel pot used for the base.  So in my version I’ve made an insulated base which looks to have solved that problem.

Here’s a video I shot going over the roaster, it also includes some clips of it running.



The base is made out of two aluminum walls fastened onto a wooden base.  I turned the base on my lathe, making a tenon around the circumference that the two walls seat against.  There is about 1″ is space between the walls.

The inside diameter is about 9.5″, enough to house the drum I’m using.  The height of the walls needs to be high enough to not only house the drum (plus a reasonable gap on the bottom for clearance) but also high enough to clear the “nose” of the turbo oven.  I forgot this second consideration and was only able to use the smaller of the two drums I bought.  The walls are long strips of 0.025″ aluminum riveted to make loops.  I used small #4 wood screws to secure them to the base.

Insulation is provided by glass fiber insulation.  It doesn’t seem to mind the heat from coffee roasting.

Bearings for the drum rod are two flanged bronze sleeve bushings, 3/8″ in inner diameter.  Because I’m using 1/4″ square rod, I needed something with a slightly bigger inner diameter to house the rod.  And nicely enough, 0.25 x 1.414 is roughly equal to 0.325.  I made sure to buy bearings with a flange to make them easier to mount in the walls.  I only then have to secure one end.

Doesn’t the wood catch on fire?

I asked myself the same question!  But no it has not, it has darkened a bit over 5+ roasts and oozed some sap.  And after each roast I can smell pine resin, but that aroma has not made its way to the beans at all.  The aluminum foil cover seems to help because it reflects the radiant heat produced by the Turbo Oven’s halogen light.

But with that said, when it comes time to re-make the base I will attempt an all-metal construction.  And I have moved a fire extinguisher to my coffee roasting station just in case!

Drum and Rod

I bought two coffee drums off of eBay for this prototype because I didn’t know what size would work the best.  Here is the listing, but in case that stops working they are each 18cm long with diameters of 14cm and 12cm.  The 12cm drum is a bit too small for a 1 pound (green weight) batch – larger, puffier beans will eventually crowd up inside and stop tumbling.  The 14cm drum has worked great on all my 1 pound batches.

The drums have a 0.6cm^2 square drive hole, which is pretty darn close to 0.25″ (6.35mm).  The rod fit in there right away without any filing.  However the holes were out of axial alignment so I had to file one of them on the corners to “rotate” it a bit to allow the shaft to slide all the way through.

Speaking of the shaft, it is a 1/4″ square solid stainless steel rod from Home Depot.  I can’t find it on their site right now or I’d provide a link.  Here is a similar product from Speedy Metals.


Roaster with old, noisy ice cream maker motor

This prototype has already seen three iterations of the motor!

Ice Cream Maker Motor

The first was off of an ice cream maker.  It worked OK but was quite loud making it difficult to hear the cracks.  (I also have not-great hearing, picking out details from background noise is difficult.)  However, one positive for this motor was that its drive shaft was designed to accept a 1/4″ square rod!  Another good thing about the motor is that it came from a $10 thrift store find so it didn’t require any eBay’ing.

4W Gear Motor

Version two has me using a 4W, 35RPM synchronous gear motor.  These things are all over eBay and about about $10; here’s one listing.  Be sure that the one you’re looking at is rated for 110V or 120V, there are also 12V models out there.

I have two of these little motors, one fixed to go clockwise and another which will randomly choose a direction when started (this is normal).  We’ll call the former CW and the latter CW/CCW, because that’s what their labels have.

In a test run with 16oz of greens the CW/CCW motor had great difficulty starting out and needed a push.  Not much more was needed to stall it, either.  The CW motor worked a lot better, still needing a push to get started but it would not stall out on its own.  As this writing I have a 14W motor on its way and will try that out and update this post when it arrives.

My first real roast with the 4W, CW motor worked really well but I only used 14oz of beans.  It stuttered a bit while starting but once going worked the whole time.  Roast #2 used 16 oz of greens and the motor would not start spinning until given a little help; but it at least ran the whole time without stopping.

14W Gear Motor

The third motor is a lot more powerful and should be the final one for this roaster.  Here’s an eBay link, but if that doesn’t work for you look for a “14W 30RPM 110V synchronous gear motor”.  It should be about $15 shipped.

While the motor’s description said it has a 6mm output shaft it’s actually 7mm so it immediately worked with the flexible coupler mentioned below.

The motor is reversible, so it has three electrical connections.  This video shows a little more detail into how to wire it up, but the gist is that one AC lead goes into the middle and the other lead goes into one of the side connections.  To change rotational direction use the opposite side.  There’s no polarity with AC.

This motor, like the 4W, is close to silent in operation but has tons of power.  It easily turned 16oz of greens, and I was unable to stop the drum with my hand.

It has four threaded bolt holes in its flange, M5-0.8.

Drive Shaft Coupling

While using a square bar for the drive shaft made it easy to mount onto the drum it made it a lot less easy to attach to the motor.

So I gripped the bar in my chuck’s pin jaws and filed the end of the square bar down to round.


Now I can couple the shaft to the motor.  The motor has a 7mm shaft, and I now have a 6.35mm shaft, so I used a flexible coupler to join the two.  I went with a flexible one because I know that none of my work is close to perfectly square or aligned.   This lets the motor and shaft rotate in peace with the flexible coupler taking out the slack.


The motor is simply screwed to a T-shaped  plywood stand.  The above photos show the 4W motor but I used the same stand for the 14W motor.


  1. Pre-heat the unit with the drum in place, slid onto the shaft and in the vessel.  Make sure drum door is pointed upward.
  2. Add beans and replace the turbo oven, set for perhaps 460F.
  3. When 1C comes around, lower the heat to around 440 in order to avoid a rolling 2C
  4. The end of the roast is still kind of awkward.  Turn off drum, remove turbo oven, and then pull out the motor and shaft leaving the drum in hand.  Now open the drum door and dump beans into cooling vessel

It’s not too different of a process as compared to my Stir Crazy, but is definitely less user-friendly when it comes to removing the beans from the device.


IMG_20180204_112727I’ve been getting very good, consistent batches out of this machine.  The drum’s agitation of the beans also does a better job of removing chaff than the stir crazy, my beans are very clean.  I did struggle with over-roasting when I used the loud ice cream maker motor because the light 2C snaps were hard to pick out.  But my last batch with the quiet synchronous motor was right on target because those first 2C snaps were easy to hear.

Stir Crazy Turbo Oven Roasting Process

I’ve had success with a very simple, cave-man like approach to roasting coffee with my SCTO setup.  Success to me means that roasts are very predictable and that I enjoy the coffee.  As an aid to other roasters that might be having trouble wit this setup I recorded a roast session.

In it I talk about pre-heating, temperature control (or lack thereof), and the cooling process.

HDPE Triple Wall Pipe for Ducting

I moved my coffee roasting station but kept my blower in the same place (less noise!) so that meant I needed to extend the fume hood’s vent line.  I went the cheap route and got a 10′ long piece of HDPE Triplewall pipe which cost about $7.60 at my local Home depot.  Its inner diameter is around 4″, and I use ducting that also has a 4″ inner diameter so an adapter was needed.  One trick is to cut off a short length of pipe, say 3″‘, and cut out enough of its circumference so that it fits inside its parental pipe.  This gives you a male end for your female duct hose.

Through the power of high school math and a pair of digital calipers I determined the required circumference removal to be 31mm.  I use metric for computed figures because applying them back to the work is easy.  My cuts/measurements weren’t perfect as you can see but I simply taped over the gap with a bit of aluminum HVAC tape.  The adapter was primed/glued in with standard PVC bonding products (i.e. purple primer and cement).

DWV Adapter Markings
DWV Adapter Markings


DWV Adapter in Place
DWV Adapter in Place

My 4″ ducting hose fit perfectly over the DIY adapter.


Clean Your Behmor!

Confession: I’ve been using a Behmor 1600 for the past year in favor of my bread machine roaster.  The primary reason was that my heat gun caught on fire, but I also wanted a drum roaster that was quiet and would let me see the beans.

So I’ve been using it for a year and have been keeping the interior walls clean like the instructions say to do.  However in the past month my roasts have been taking a long time, 10oz of beans pushing even the boundaries of the 1# setting.  Flipping through the manual alerted to cleaning the inside of the machine as well, and I found the cause of my problems.

Inside view of air intake fan
Outside view of air intake fan. Yuck.
Exhaust chamber

I cleaned out the dust & lint with my vacuum’s angled attachment and roast times went back to normal.  My best guess is that because it was not able to draw enough air to keep the electronics cool so the roaster lessened its heat output as to not cook itself to death.

A word of advice though, after you get the side panel off plug the roaster back in before vacuuming it out in order to ground it and to dissipate any static electricity buildup.  Otherwise you may fry the circuit board as vacuums generate a lot of static.  Just be careful about poking your fingers in there!

Speedy Roasts

A problem I’ve had with my chaff collector hood is that it retains too much heat thus allowing the beans to heat up faster than they ought to.  When I set the PID to 400 and let ‘er rip I get roasts that look like this:



Notice the mottled color, uneven roasting, and the fissures.  The coffee is bitter with no depth.

Edit: I’ve recently learned that the effects of over-heating the beans is called “tipping”, because the small ends of the bean tend to get burned first.

Tonight I did a batch following a more gentle curve and got a much better-looking roast:

They already smell better than the bad roast, and they’re 10 minutes old.  I managed to keep the temperature right inline with this recommended roast profile by setting each minute’s taget temperature at the start of the previous minute.  For example if at 4 minutes I was suppoesd to be at 180 I would start adjusting my PID at 2:50 which takes about 10 seconds to increase by 20 degrees.  Then over the next minute the PID will gradually bring it up to the correct temp.  They key here is that when the PID is close to its target temperature it won’t go full power.  If you’re 300 degrees away from your target it’s going to apply heat 100% of the time which apparently leads to crappy roasts like you see here.

Coffee Roaster Chaff Collector

Having an open top bread machine coffee roaster has two drawbacks, the first being that chaff flies all over your workspace with the second being there’s a lot of wasted heat.  I made a lid/exhaust vent/chaff collector in an attempt to solve these problems.

My bread machine came with a lid for the bread pan which would be used for its rice cooking mode.  I put two holes in this, one for the heat gun and the other for the exhaust pipe.  The pipe is a piece of galvanized ducting rolled into a tube, secured with pop rivets.  On one end I cut a series of slits to make tabs, these were used to pop-rivet the pipe onto the lid.  The tube is 2.25″ in diameter.

For the flexible tubing I found some 3″ corrugated aluminum ducting at Lowe’s.

The tube/duct join needs some cleaning up!  The end of the heat gun snout is slightly smaller that the rest so it nicely rests in the lid.  You can also see how the plastic body of the heat gun is melted, this is from when I would simply stick the heat gun into the pan.  With the lid in place the heat gun won’t melt anymore and there’s no chaff to get sucked into the gun!

Some of the chaff collected, about halfway through a roast.  There was a lot more at the end.

The most obvious drawbacks of this setup is that one can’t see the beans and they’re more difficult to hear.  I do plan on making a window on top with a small piece of tempered glass but the sound part is harder to tackle.  1st crack is audiable but I’m pretty sure 2nd crack is totally muffled.

Cory Pulldowns, Fixed

For a long while I was having a terrible time with stuck “pulldowns” aka “the trip south” with my Cory vacuum pot.  My seal was good but the coffee would start with a trickle and then turn into frothy foam, and finally stop alltogether.  The amount of vacuum was incredible as the glass rod would only turn with some difficulty – I’m luck I never had an imploded carafe!  I tried the “new” cory rod but it didn’t perform any better.  I then used a my Demel’s diamond cutoff wheel to make a bunch of notches in the new rod but even that didn’t help!  The most consistent solution I found was to tilt the rod right before pulldown in order to create a larger gap.  This would get the pulldown phase down to 5 minutes, still pretty bad.  And of course it would still get stuck about half the time forcing me to filter the remainder of the slurry with my french press.

It turns out my problem was with how I was whirlpooling the coffee as I turned the heat off.  I liked how this method would result in a nice cone of grounds but for me it also packs the gap between the rod and bowl with fine ground dust.  This dust would form kind of a mortar resulting in stuck pulldowns.  I’ve stopped whirlpooling and my pulldowns, with my original rod, take about 30-60 seconds now.

I’m guessing that the whirlpool forces the smaller particles down to the bottom before the larger whole grounds.  I do notice now that after extraction the dust forms sort of a slime on top of the grounds wheras with whirlpooling the dust wasn’t noticable meaning it was more evenly distributed, or perhaps mostly on the bottom near the filter.  I have a low-end burr grinder which produces a lot of dust.  Those who sucessfully whirlpool likely use much more consistent grinders.

Bread Machine Coffee Roaster Version 2

Update: This post has been converted into an updated, static page.

I’ve made my coffee roaster more self contained and easier to use by:

  • Moving the PID controller and solid state relay (SSR) into the unit
  • Mounting the thermocouple through the bread pan wall
  • Shortening the mixer arm
  • Separating the control of the heat gun’s fan and heating element

What’s in the Box?

Here’s a schematic diagram of the roaster,

Electrical Schematic
Electrical Schematic

and here’s a photo of the exterior.

Machine showing control panel and thermocouple plugs
Machine showing control panel and thermocouple plugs

I cut out the middle of the old control panel and bolted on a piece of aluminum to serve as the new faceplate.  To this I added the PID controller, main on/off switch, mixer switch, and the heat gun switch.  On the left side of the unit the thermocouple banana jacks are visible.

Back of Control Panel
Back of Control Panel

To make room for the SSR and heatsink I removed bread machine’s ventilation squirrel cage fan.

Cover removed shoing SSR, controls, and mixing motor
Cover removed showing SSR, controls, and mixing motor

When using a SSR it’s a good idea to either mount it to a thick piece of metal or to use a heatsink – they get pretty warm.  If you look closely you can see how its label has peeled up and warped; this is from when I ran the SSR bare on a different project.

SSR and heatsink.
SSR and heatsink.

The wall outlet is where I plug in the heat gun.  The wiring on this is really bad/unsafe.   I used the ground connector for the AC common and the main contacts for separately powering the heatgun fan & heating element.  If anyone ever plugs the heat gun into a normal outlet I think the element would turn on without the fan thus burning it out.  Also this would be putting about 10A through the ground conduit which is probably bad.  At the big box hardware stores I looked for custom 3-contact plugs but I couldn’t find any – my next McMaster order will include something like this.

Side of machine showing heat gun plug
Side of machine showing heat gun plug

I used a bus bar to try and keep the wiring tidy.  A note about wire gauges: I used 14 gauge stranded for everything leading into the heating element because it should draw about 10 amps.  otherwise I used 18 gauge stranded.  The DC stuff used 22 gauge solid core.

Wiring bus board used for splits
Wiring bus board used for splits

Fun fact: the heat gun switch (far right) is the switch off the heat gun.

Closeup of the control panel: PID, on/off, mixer, and heat gun.
Closeup of the control panel: PID, on/off, mixer, and heat gun.


I got a bayonet-style thermocouple because it’s easily removable from the pan.

Bayonet style thermocouple
Bayonet style thermocouple
Exterior view of bayonet mount
Exterior view of bayonet mount

Here’s the inside of the pan with 8oz of beans.  This is the lowest I could mount the thermocouple without the probe hitting the mixing arm.  While it’s not immersed in the beans at rest, once they’re agitated they go over the top if it just fine.  In the end I feel that I’m getting a good, consistent read on the external temperature of my beans.

8oz of beans in the hopper
8oz of beans in the hopper

Heat Gun

I’m using a $10 Harbor Freight heat gun with separate power control for the fan and the heating element.  I separated them because I felt rapidly cycling the fan motor on/off would shorten its life.  So the motor is directly powered by heat gun switch while the heater has the SSR acting as a secondary switch on top of the heat gun switch.  I always run the heat gun in “high” mode because the low mode wouldn’t heat fast enough.

As mentioned above I moved the heat gun’s switch to the control panel .  Within the heat gun I simply wired the two power leads to the fan motor and element.

Mixer Arm

I cut off about 1/8″ of the mixer arm because its normal length was causing excessive bean binding between the probe and the arm.  Now when it runs I don’t hear any the loud pops caused by beans getting jammed.

The Mixer arm with 1/8" taken off
The Mixer arm with 1/8


To roast I simply stick the barrel of the heat gun into the pan as far it will go and let ‘er rip.  The thermocouple is in the top left corner of the pan while the gun is pointed in the bottom left.  I do keep the position of the gun the same for every roast so that the the pan heats the same way.  In this photo I’m experimenting with an advanced heat retention system.

How I hold the heat gun in the pan
How I hold the heat gun in the pan

Parts List

  • Hitachi HB-201 breadmaker from Goodwill
  • HarborFreight heat gun.  It’s not on their website but this is the $10 model with only a high and low setting
  • SET-620 PID controller, purchased off of eBay for about $35.  This is a made-in-China el cheapo unit but it works OK
  • 2 meter K-type bayonet thermocouple purchased from “Procon Products” on eBay
  • M12 x 1.5 hex nut used to secure the thermocouple mount.  I was able to get this at my local old-school Ace Hardware
  • The on/off switch is a beefy 20A-rated SPST toggle switch
  • The mixer switch is a DPDT switch leftover from another project.  I think it’s rated around 8A.
  • The SSR is another made-in-China eBay aquisition.  Do a search for “SSR heatsink” and you’ll see a bunch of listings for SSRs with matching heatsinks.


I’m pretty happy with this setup.  If anything it seems to be a bit underpowered but this could be alleviated by making a lid to retain heat (Updated).  I like it because I can keep an ear and an eye on the beans all for a pretty minimal investment.

First Coffee Roasting Attempt

Tonight I roasted my first batch of coffee beans.  They’re darker than I would have liked but I think they’ll be palatable.

My roaster is a bread maker and hot air gun.  I hot-wired the bread machine’s stirring motor to a switch, and for now I simply point the air gun down into the basket.  The dough hook does a great job of keeping the beans agitated so that they evenly roast.  The red toolbox is my temperature controller.  Today I only used it as a thermometer but soon I’ll install a thermocouple inside the bread pan.

The Roasting Setup

Here’s a closer shot of the thermocouple and my advanced mounting method.  It’s positioned such that the tip of the probe is down inside the beans.  It never showed any hint of wanting to come loose so this could be a permanent solution but I’m a tinkerer so I’ll make something more airtight later on.

Temperature Probe

An important step in roasting is rapid cooling after you’ve hit your desired level of roast.  I’m using a pizza pan & fan which worked out just fine.

Cooling Aparatus
Cooling Aparatus

The roast itself went pretty well except it got too hot right after the first crack and I think the second crack happened too soon for me to notice.  I was going to stop the roast when the second crack started but because I hadn’t heard a second set of cracks I kept at it for a bit more and wound up over-roasting the beans.  They looked good towards the end of the cracking, i.e. they had a nice brown color and weren’t too oily; I should have stopped there.

Mine are on the top right
Mine are on the top-right

In this photo my beans are in the top-right canister – too dark as you can see.  The other canister is Alterra’s Black & Tan blend, even its dark beans aren’t as dark as mine.  The bag has the green beans.

Roasting is a smokey affair, especially when you over-do it!  But overall it’s a straightforward process and should be easy enough to get proficient at.  It’s just like baking but in a 8 minutes.

Don't do this indoors