Rob, I hope you don't mind that I consolidate your comments and pictures into one place... I go looking for them all the time and thought we need one resource. Please comment and update at your leisure.
I have built several air roasters. Started with a 650 gram roaster and worked out we could use a bigger roast chamber on this unit that took it to 2.3kg. Made a slightly larger Rc and managed to get the max load to 3kg. Blower is a 230V 1000watt vacuum cleaner motor, heat source is a 2941 sievert LPG torch burner. Currently nearing completetion of a 5kg RC for same roaster will have 2 Vacuum blowers and probably a 2942 gas burner. I would go for the vacuum blower they fall in the right range for pressure and flow. The 2.3 kg RC is 6" dia and bean bed depth is 8". RC is round with assymetrical bean flow. IMO works better than the cone type RC, appears to need less pressure to get and keep the beans moving. There are about 5 roasters like this out there with some of them having done over 1000 roasts and no problems
my first gas air roaster attemptis on Ed Needhams site I must be onto the mark 7 or something by now.
The LPG torch sitsin the middle of a 6" dia pipe, i have a disc approx 4.5" behind it to stop the blower blowing directly at the flame. The first roaster had the blower at the bottm of the sam e pipe later units have the blower in a separate pipe with RHS directing the air in from under (pics will make it clear) My roast chambers are round but the bean flow is assymmetrical. That is - not a cone shaped roast chamber. My 10" RC is conical shape but my experiments point towards the assymetrical chambers requiring less pressure and flow to get the beans moving. I think this is because there is less friction with the beans to the side of the stainless steel pipe compared to the cone type RC having to deal with the bean to bean friction
This is a picture of the mark 2. I am still using the furnace section of this. Th RC is the 650 gram unit. Have progressively built a 1.9kg, a 2.3kg and a 3kg. Working on the 5kg at present.
No lid required (although the stainless one has a perf lid for when you dump the beans, long story). I have a triac based 1kw controller for controlling the airflow and hence bean movement and also the profile. Temp is measured in the none airflow part of the bean mass. Dump temp is typically 233C which is early part of rolling 2nd, a few beans have oil spots. Roast starts with controller at bout 80% and finishes with it at bout 40%.
The beans flow very smoothly across the roast chamber, this is an assymetrical type RC (haven't used the cone one yet). Blower is a 1.1kw cheapo vacuum cleaner motor bout 5" dia and only 4" high. We have fluidised 3kg with one blower.
Typical roast is 1st crack at bout 8min and 2nd bout 11.30. Roaster is very controllable profile wise. Am working on a temp display that gives more data. Basically a datalogger that displays delta T, ie rate of temp increase per minute. had a friend make one at one stage but cos we used the wrong A to D chips it awsn't as accurate as we had hoped. There are some guys around the cnr from my work that have a datalogger they can change some code for a moderate charge and it should do what i want.
The vacuum cleaner motors i have been using are the 1100watt (230v) cheapo units. Tested a 1600watt one, found it had no more blow than the 1100 unit. Someone (here i think) had a link to the Graingers catalogue that had a link to the manafacturers spec sheets. Look for the max pressure (same as it's vacuum) and max flow without getting into too high a wattage unit. 1100watts at 110volts will do the same work as at 230volts.
The LPG (Liqiud petroleum gas) torches draw in air at the torch head. The 2941 unit is rated at 7KW more than enough to roast up to 3KG (or more) of coffee. There is a 2940 that is 3 KW (used it for a few roasts with the 650gram RC way back when). There is also a 2942 that is 26KWs. If you need more than that you are trying to roast too much coffee too quickly.
Third item down Art no: 219741 is the torch i use, also with a 1 to 4 bar pressure reducer. Price here runs to bout NZ$150. I also swop the neck tube for a longer one that i cut and braze thru the bottom plate of the roaster to a turned and drilled piece of round bar.
The first roaster (the one on Ed's site) had the tube set between two 6" pipes and sealed with tape ( hey it worked).
My roasters use only 25% or less of the roaster diameter perforated. It wouldn't work with all of a 6" base plate perforated but that IMO would be an inefficiant way of doing it. I just don't see a plus side to perforating all of the bottom plate.
I must try to do a video on UTube of the roaster in action, the beans flow really smoothly across the RC.
View of the under side, brass neck tube is cut bent and end braised into round bar which is welded thru the base plate:
Top View with "air deflector plate":
View of burner, electric ignition in testing stages jury rigged onto this unit. Vacuum cleaner motor and mounting plate in view to the right. I once made a bracket like this by hand. Cut by hand with the oxy-acetylene torch used a cutting disc for the hole, took hour & a half to do. Same bracket from a profile cutting place cost $6, (in conjunction with a batch of 20 other plates discs etc):
burner specs below
3940 (i called it a 2940 in previous reply
3 kilowatt or 10600 BTU/hr
2941 (the one i have used, good for from 650grams to 3kg+)
7kw or 26300 BTU/hr
26 kw or 88700 BTU/hr
The 5 kg RC might need the 2942 (it will in winter thats for sure).
A better product catalogue than on the earlier post.
Hi Lynn, yup thats a drum roaster which i will be getting stuck into as soon as the 5kg setup is finished.
Only that one stainlees steel roaster had a dump setup. For dumping the 45 deg plate was tilted up to vertical, the beans then dropped behind it and out of the shute instead of building up and flowing back down into the air flow. This unit had a perf plate that was placed on the top of the RC for dumping, cos as the beans got blown out the remaining beans started to blow out the top.
The vacuum blower sucks air thru the cooling bin to cool the beans. . Works good with a 650gram RC, but with this SS (2kg +/-) RC the motor last about 15 months of moderate use. I kinda expected this. I have built a cooling bin for my own roaster out of a wood shop dust collector.
One of these: http://www.topmaq...index.aspx
It's a bit Heath Robinson, but that is a futue project to upgrade.
Cooling beans by blowing air thru the RC would take too long as there is a lot of heat left in the steel of the RC, my guess would be well over 5 minutes.
The conical RC works well, i roasted 4kg with it using 1 only 1000watt VC blower and the 7kw sievert burner. I know I can move 6kgs in it using 2 blowers. The reason i went for the conical RC was i tested an 8" asymmetrical RC with 2 blowers and could not get much more than 3kg moving well in it. Also the beans fell back in the middle of the pile not across the other side of the RC. Because of the perceived problems with this RC I kept the perfed area of the conical Rc small - less than 2" dia.
Today i revisited the 8" asymmetrical RC. I redrilled the holes from 3mm to 3.3mm, broke the drill half way thru and finished the last half with a 3.2mm drill bit. Tested it cold with the 2 blower roaster that is almost complete. I had 6 kg of green dancing right up to the edge of the RC. Could probably do 7kg with it except once roasted it would be blowing out the top. Could look at a curved deflector to sort this.
The really good thing about this RC is that by just turning down the dimmer/controller on the VC blower the beans are roasted at a reasonable roast profile.
The height of the bean mass is limited by the airpressurre available and the how the beans behave once roasted. Any more than 2.3kg in my 6" Rc and I have roasted beans bouncing out the top toward the end of the roast, again a deflector at the top would probably mean i could do closer to 3kg. My earlier 6" RC would only do 1.9kg of green because it was shorter.
Problem now with 2 blowers and still only a 6" dia "furnace" section of the roaster is that the extra airflow snuffs the burner out. Something i need to work on
CFM is volume. You need pressure, the VC cleaner blowers have plenty of that. From memory I measured one at about 18" water, the blower at the top of the page would be lucky to have 2". Even a small VC blower would be good for a kg of coffee even more.
If you read some of the earlier threads about blowers and air roasters there is alot of info there about this.
Initially I came to the conclussion that you would need to go conical over 6" dia. But after getting 6kg lofting in an 8" asymmetrical RC I don't know if there is an upper limit.
I do believe there is an advantage using the asymmetrical design in minimizing bean to bean friction. I tried something random the other day. had 5kg of green in the conical RC, it wouldn't loft with the blower on full noise so i pushed a 2" ss pipe down the middle of the RC to about an inch or so from the perf plate, the beans started to flow up and spout out of the pipe. Seems the pipe directed the air flow and cut down the friction/engulfment of the surrounding beans.
The perforated area should be related to the total area of the RC say 20%. Hole size and distribution is linked to the pressure and air flow. Too many holes for a given area and the you lose pressure & velocity and get a random bean flow and bean bed height is limited. If the holes are too small (or too few for a given area) again your bean bed height will be limited. Greater pressure can over come this but is probably not an option due to limited options for blowers. Also heat/air density will no doubt be affected.
Your bean height is affected by air pressure. If your air pressure is 18" of water, beans weigh approx 70% of water so you should be able lift more than an 18" column of beans. However because the beans don't form a tight seal in a pipe you get alot of slippage (airflowing past the beans). Plus you need surplus airflow to throw the beans around.
So my ball park bean bed height is around 8" give or take.
The bean bed depth is about 8" before it goes on the roaster once on the roaster the riser is blowing bout 8" above the bed. Toward the end of the roast it can be just below the top of the RC which is 26" from the perf plate.
VBM Minimax 2gp, 1gp Reneka Techno, 2 gp la Pavoni Pub, la Cimbali M28, SJ Maz, FB 6kg HM roaster and other stuff
. really dig the simplicty of these machines. A few questions. .what size is the orifice coming out of the blower. how much distance is there between the blower and the heater. What temperture are you getting pre RC chamber (if your monitoring) and what temp in the RC chamber. Lastly what technology are you using to monitor heat. cheers!
Yes, but if airflow could be turned down to keep beans a foot lower... and heat would have to be turned down too.
Rob beat me to it, I definitely meant to say that airflow needs to be adjusted during the roast. I'm shopping for a borosilicate glass pipe for the next iteration so I can keep circulation dialed in without having to peer down the shoot. Any leads? The local hydroponics supply is my best source yet.
I've used Rob's design to roast over six hundred pounds in the last six months. I'm finding a lot of transparency in the roast process using his design. Despite the singular drawback of airflow being tied to heat, I don't think I'd be interested in another roasting system. I can get exactly what I want out of the bean, and I don't think a drum could give me this level of control given it's closed environment and thermal mass.
I'ved used Robs design and have at least 26 empty coffee sacks in the garage ;)
Blower is a from a VC no single round hole, 2 splits (from memory) where the air blows thru the motor. You would have to look at one.
On this unit blower is in the part that has all the masking tape blocking the 2nd blower see still above of youtube clip.
Air goes down SHS section to bottom of 'furnace' section.
Comes out of hole below burner see photo above dated 28/07/2005. Photo above that shows plate that protects air flow from burner. I gave up monitoring temp just below RC. I can profile based on actual temp increase per 30sec. Just get the start flow and burner setting right and adjust as you go.
From about 3mins I am turning the dimmer controller down about a 1mm every 30secs.
Original roaster had the blower about 6 or 8" below the burner see Ed Needhams website for the original prototype.
I use a Zen PID - Be the PID- I am the PID.
Very simple TC display, I manually log temp every 30secs (bean mass temp out side of the main air flow) and adjust accordingly. Very easy to profile once you get the hang of it.
There are at least 5 of these roasters with collectively a minimum of 7000 roast between them.
VBM Minimax 2gp, 1gp Reneka Techno, 2 gp la Pavoni Pub, la Cimbali M28, SJ Maz, FB 6kg HM roaster and other stuff
yeah, thats a great pric. The numbers i had quoted from gray glass where 3 times that per foot.. but that was 10" and its only made in medium wall..I didnt get pricing on the 6" 0r 8" but you can get it in standard wall. which should be considerably cheaper and they will cut to order.. look them up ..
Some video of my roaster at the start and finish of the roast.
OK video production not my strong point.
Dumping pretty quick. I have a 3/4hp woodshop extractor blower under the cooling bin- beans cold in about 3min. Cooling bin is about 16" dia. New bin will be bout 32" dia.
Great to see some more videos. I just added a 90 degree elbow to mine to help blow chaff out the window. I've also been playing with a damper to retain heat. The heat gun in my version is such a limiting factor that I'm going to have to move to gas soon. I'm still trying to work out an apartment friendly/indoor solution for that.
I like the RC handle/thermometer brace. I had not come up with a good way to mount the thermometer permanently.