I've updated this thread here, rather than adding another post.

Version 0.1
I'm in the prototype phase of making a Milk Can Roaster. The milk can is stainless (5 liter). I made a stainless shaft, coupler and stainless perfed vane on the inside. I've only done two roasts using a heatgun as the heat source. It does 1.5 pounds nicely. Very even roast, no burnt beans, tipping etc.



Once I roast another 20 or so batches I'll use real bearings and a motor. Right now I'm getting the feel of it. I may also change to a propane stove burner underneath for all or most of the heat. I don't know how well that will work. [Edit: I like my wood bearings just fine! The propane burner works very well! - see video below.]



The inside vane is tilted so that the beans in the back are spilled towards the front and the unit itself is tilted so the beans in the front slide to the back.



Any suggestions would be welcomed. Entire budget for this will be less than $100. I was thinking I'd have to punch some holes in the side, but that doesn't seem necessary at the moment.

Update:

I've switched to propane as the heat source, but I haven't had enough experience with the machine to do a good job of controlling the roast. With the heat gun, everything was smooth and controllable. With the propane burner it wants to get hot too fast. I'll do a few more batches and try to creep up on it a bit.

[Update] It is easy to control the heat/ramp/profile. I just needed a couple batches to get used to it.

I also got tired of cranking the drum by hand. Couldn't keep notes and watch the roast etc. at the same time. So, I hooked the drum up to my Tormek knife sharpener. It turns at 90 revs/minute. Once I get the temperature dialed in better I'll have time to sharpen a knife or two before it gets to first crack...(yes, I'm serious). I could be wrong, but this might be the first time in the history of coffee that a roaster and knife sharpener were combined.

Here is a video of the roaster in action.

http://www.youtube.com/watch?v=QBSeh69h2_I&feature=channel

Here is a video of the cooler I have, which is actually incredible and useful for all roasters.

http://www.youtube.com/watch?v=S2mRgPpIFxA&feature=channel

Yes, I have sharpened knives while roasting. It takes about 2 or 3 minutes to sharpen a knife razor sharp, and I'll do it during the first few minutes of the roast. After that, no way!



dan

Nice work. It appears to have great craftsmenship. If it were me, I would probably add another agitation fin but I haven't seen it in action either.

If you do decide to use a gas burner for a heat source, I would suggest some sort of can/cylinder to focus the heat on your milk-can/drum.

check this video of the infamous SCR (soup can roaster): http://video.google.com/videoplay?docid...897947291#

You might google the SCR as well to help gather ideas as it appears your concept parallels it closely.

It looks as though you could add a cheap AC motor to automate the drum rotation which might allow you to watch the roast more closely. Then again, I really like the hands on approach that you have - it helps capture the essence of home roasting.

I'm sure there will be a few folks out there that would be tempted to follow in your foot-steps as well so a materials list and "how-to" guide would probably be much appreciated.

Also - curious what your roast times are in that bad boy? How long till' you hit first crack?

keep us updated - that's a very cool project!

cheers,
j

Third Roast Data

Suggestions/comments welcome. I've done about 300 roasts using the heatgun/dogbowl method and this is my first foray into making a roaster. I believe that the roaster would handle an additional 16 oz (by volume) pretty comfortably.

Quantity : 32 OZ by volume (one large yogurt container)

Cold roaster to beginning of 1st Crack 13 Minutes 37 Seconds
Beginning of 1st Crack to End of 1st Crack 2 Minutes 11 Seconds [u](Is this good?)[/u]
Additional Roast time, almost beginning 2d Crack, but not, 1.5 minutes (Should I slow it down a bit here?)

Cooling using my Dust Collector 16 Seconds from dump to ambient temperature!

Very even nice looking roast. The chaff tends to blow out during the roast. Turned crank at about 30 revs per minute.

This seems promising. I'll do a few more and make a few tweaks before motorizing it.

dan

Switched Heat Source

For the last few roasts I've switched to a propane burner (outdoor stove) as a heat source. There is a learning curve. The heat gun allows more control quickly. However it was taking too long to get to first crack and I was tasting some "baked" tastes which I didn't like. The propane burner can get me there way too fast. I'm using an IR thermometer to take some interior measurements, but it is hard to turn the drum and measure etc at the same time. It will take a few more roasts to get the thing under control. But the taste seems a bit different than the heat gun. I wonder if the volatile gasses etc are removed by the heatgun and are kept by the propane method. I'm not sure yet, but I think the taste is more rounded and integrated as opposed to more edgy and easy to pick out flavors. I wonder if this makes sense?

Bottom line: Very tasty roasts, but not as even and pretty as the heat gun. This is because I need to ramp the heat a bit slower initially. If anyone is interested, I'll post more pictures.

BTW, the dust collector unit I use for cooling is simply spectacular. A colander of smoking beans brought to room temp in 10-15 seconds!

Update
I roasted a couple batches (full 32 oz yogurt containers) this morning. I let the heat start out low, gradually increasing it after the beans were in the yellow grassy state (about 3 minutes). I fiddled with the heat along the way, reducing it at 1st crack which was about 10 minutes into the roast.

One batch I roasted to a solid 2d, the other I didn't let get to 2d. Both look very nice.

I made it much easier by using a stainless tube which I cut the tip off (making it half round). It is easy to grab a batch of beans anytime and dump them back after examining them.

I'm going to do a few more tweaks and then make a video after I'm more confident in controlling the burner. Anyhow, it doesn't allow the complete precision that heat gun/bowl allows, but does do a very even, nice roast of 32 ounces at time.

dan

Knife Sharpener Update

I made a minor mod to the roaster (made the hole in the lid smaller) and roasted 32 ounces of an Ethiopian dp. I purposely started low and after five minutes, increased the heat until the first of first cracks, after which I lowered it and let it crack for about 2.5 minutes. At this point I was 10 minutes into the roast. I let it go for another 4 minutes and before 2d crack, cooled it. Elapsed time = 14 minutes. I may take it a bit slower tomorrow with the goal of ending at 16 minutes.

I used an IR gun style thermometer which gave me indications of temperature and temp rise. I didn't let it go above about 410 f. I could easily have done so, however. The thermometer is not a very good way of measuring-- it gives a rough indication.

Very even, beautiful roast. No tipping, no visible defects. I made some drip to taste it and it was even an hour after roasting, very nice. I'll do a few more and then make a short video. Before I started this thread I did more than 200 roasts with the heat gun/bowl method which was great training to know how the beans are developing with the roast. If I hadn't done that, I don't think I'd be having a successful time with the propane powered roaster. The heatgun method, however, won't do 32 ounces evenly; 16 ounces was as far as I got.

Very satisfying to dump the beans into a 32 ounce Mason jar and have a small bag left over to evaluate and/or give away.

dan

Added a couple videos and updated the original post.

dan

Congratulation.

I have been following your endeavours with great interest. I am happy to see that the roast are coming out nicely and that you are persisting.

I also have trying to built a roaster from scratch but ran into all sorts of problems. I was too ambitious and didn't know enough... but I am learning and will post when I have a working design.

I have one suggestion for you, do with it as you please, I think that having an enclosure around the can , however crude it may be, would probably help a lot in controlling and distributing the heat.

One question, just curiosity, how do you empty the roaster?

Please continue posting.

Cheers,
Sylvain

I think I may make some shrouding for the roaster, though it doesn't seem to need it. It might be more energy efficient, though, so I'll rig up something.

I will post a video of emptying the roaster. Basically, I take the drum off (it has a coupling that allows this) and with my right hand remove the top using my stainless steel pipe/trier. Then I dump the beans into a collander and onto the dust collector where they are cooled in 14 seconds. I need two hands, so I need to get someone to make the video... In the meantime, here's a shot I did tonight 48 hours after roasting.



This roaster was pretty easy to make, and works very well. I wish the beans didn't make so much noise as it turns because with some coffees (not all) it is hard to hear the beginning of second cracks. But if I pay attention, it isn't a problem. I thought of using cast iron which would be quieter, but it stores too much heat and would be much less responsive to temperature control. Sometimes I use my trier as a sort of stethoscope which does allow me to hear inside the drum. I can also puff a bit of air into the drum which causes smoke to come out for a sniff.

I roast about 64 ounces (dry measure--two large yogurt containers) a week or so, and I can't imagine dialing in a profile for repeatability, since I only buy beans in five or ten pound batches. What I tend to do is on the first batch do as light a roast as I can get away with and still have a nice drip. I evaluate the coffee and the next time I roast, I go as dark as I can without finishing 2d crack. Now I have the two extremes. Between them I can play around with the temperature, for example, decreasing the temperature after 1st crack and letting it coast for a couple minutes or so.

By the time I've finished trying out this and that, the ten pounds is used up and before that, I've bought another five or ten pound bag or two.

dan