Jim, any thoughts on why the Sivetz air roaster isn't used more by the top micro-roasters? I have a local roaster that has used one for many years but it's hard for me to really evaluate the roasts as they tend to roast their beans really dark.
Now that you've sampled some roasts with the Quest and have decent control with it will it be your roaster of choice and what role will your popper have?

I'm not sure yet.

So far I don't get any real differences on the origin flavors or aromatics. For very fast clean roasts, my air roaster is easier, since it runs cooler. But that is more about the insulation on the two particular roasters

On the drum, I can flirt with the malty, bready and woody Maillard flavors that develop around 350F, and instead of flattening and ruining the roast, I get a more complex and interesting roast flavors. I suspect this could be a fundamental difference between lower and higher air flow roasting. I took a cupping course where they used the little Probats and noted this difference; and it's held up here. But I can't really manipulate it yet; so I'm not sure where the exact cause lies.

One very good thing about the M3 is that it has a roasting wider range than my air set up. Even with variable airflow, I can only roast from a 120 to a 180 gram range on the poppers. On the M3, I can run 12 to 14 minute espresso roasts with loads from 100 to 250 grams, and 10 to 11 minute cupping roasts from 100 to 175 grams. This gives me a very welcome flexibility when I'm using up coffees at different rates. I suppose this is true of all drum roasters.

The M3 appears to be quite the roaster!

I've been watching this thread with great interest since I've been experimenting for some time with a multi-heat source home build ( solid drum with individual ramp/soak PID control of convection and inside the drum radiant).
I've gotten many good roasts and quite a few so-so utilizing a few of the posted curves (quick to 260..drying phase...slow from C1 through end of roast) and many variations on this theme.

Something I've been curious about for years but have not seen posted is information showing preferred shop roaster (drum) curves for high and low grown beans. By shop roaster I'm thinking along the line of the Probat L-12 size roaster.

I'm sure that most folks making their living roasting on these aren't dying to give away their trade secrets and is probably why the scarcity of information. I also understand that these roasters can't turn on a dime like our home roasters so they're probably not very complicated curves to begin with but I've had some of the most complex, wine-like coffees roasted through these.

It would be an interesting comparison to cup a varietal roasted in a 12k shop roaster (by a seasoned pro) and using the same curve through the M3.

On a final note, I roasted professionally on a one-bag Sivetz for a while and I think the reason for the lack of wide spread use of these is their finicky nature and lack of sex appeal. We were always having to tweak the guy to keep a consistent bean lift and since most folks use them without ramp-soak capabilities, they have difficulty controlling the roast via gas pressure manipulation. Others just go full throttle killing the roast at a specific temperature.

Anyway, love the home roasting posts I've read so far and look forward to all the future ones.

Allen

After further thought about my above post I realised I made the typical newbie error of injecting a lot of off-topic items. My appologies for this and I'll refraiin in future posts.

Can the admin/moderator pull my post and I can start again as a new thread? I would hate for this to pull the thread down a side road.

Thanks,

Allen

AllenB, you look alright to me.

This is a test roast where I have put a towel around the intake hole area (see Exhaust for the Quest M3 roaster). This has caused several things to happen. The intake air is now more clean (filtered) in terms of variations of temperature. Now the MET does not suffer from pockets of cold air and looks more constant. It has increased the air intake temperature to the 130F (room temp = 67F) and increased the heat transfered. And I think it has lowered the power consumption of the roaster (PID stays off longer). The times have shortened also.

This graph is the same beans as the previous roasts. Also 150 grams. It started at lower temp (197F) and it reached the start of 1C faster. No adjustments made.



Cheers

allenb wrote:After further thought about my above post I realised I made the typical newbie error of injecting a lot of off-topic items.

It's fine. The problem is that aside from the Willem Boot articles in Roast Magazine, which are available on line as PDFs, there is not much on profiling high and low growns. His interesting point is that he thinks even the drop in needs to be lower for low growns. I wonder if this has less to do with their low desnity and more with higher moisture content? Or maybe the carried over heat from drop in to the rest of the roast is a lot more on a shop roaster size than a sample roast one?

I've done a lot of profile matching experiments over the years. Matching just the bean profile will not get you the same taste (close, but not quite), you also have to match the environmental temperatures. Even with that, I get subtle difference with the malty, woody, grainy flavors that develop early in the roast -- there's always more of them on the little M3 drum than on my little air roaster.

I have no real handle on why this difference develops. My best clue so far is that the timing of the early roast is much more finicky in the air roaster; with about a two minute window from the start to first crack for acceptable roasts. The drum, on the other hand, has about four to five minutes of wriggle room for the same period. This allows me tune these flavors more easily. Since they are new to my own roasts, I'm probably overdoing them.

Thanks Jim and Rafael.

Jim,

You've given food for thought on bean density/drop-in temps. On the Sivetz we always had to ramp up less aggressively when roasting Brazils and other lesser dense beans during the first 3-4 min or else we'd always have developement issues causing a flattened cup or other off-notes.

Thanks again,

Allen

Jim,

I have but two questions: number one is, how useful is the original thermometer of the M3 and what is relation of the temperature it showes to the temperatures you measure? And the second question is more related to the M3 roaster itself: could you elaborate a bit more on built quality, maintenance, cleaning, how it is working after the few months that passed since you wrote the initial imopresions?
This thread is very informative on many subjects within roasting topics (big thank you) but with regard to the roaster itself there is not said much.

Regards,
Pawel

Since I received it, around Thanksgiving, I've logged fifty odd roasts.

Maintenance: Except for the infant mortality burn out of the heat element (Tom had it too, they were supplied a bad lot of heating elements), the unit has been bulletproof so far. They handled the repair by express mailing two replacement heaters (in case both were bad); which was confidence inspiring. The drum and roast chamber are easy to clean. The exhaust components get lined with a combination of chaff dust and a resinous/oily binder. So the tubes and air box need to be disassembled, scrubbed in water and dried to be clean. It started feeling funky to the touch after 25 roasts, but there was no degradation in airflow or roast speed before I cleaned it after about 40 roasts. The clean and scrub took about a half hour. The design is really maintenance friendly, so it's almost more pleasure than chore to keep it in trim.

Operation: So far I'm enjoying the all-manual experience. A roast session works as follows. Heat up the roaster for about 20 minutes; reset the 45 minute timer switch to full time at each roast start*, drop in the beans, adjust the air and heat every minute. Using the bean thermometer (which is very accurate and repeatable), you'll see a response around 90 seconds later, so I recommend adding a probe through the top left screw to monitor drum temperature, since changes show up there in about 30 seconds. To end the roast, in addition to temperatures, you can use the tryer, the window, or sniff the airflow. I'm an air sniffer, but if Steve is right in his comments on the HB roast competition, I'm jumping the gun by about 30 seconds. Once the roast is dumped by opening the door, the heat is turned all the way down, the air is turnd all the way up, the cooling tray is transferred to the air box at the top rear of the roaster. When it is slid in, it blocks the air out of the drum, and draws cool air throiugh the beans. You need to stir manually for about 2 to 3 minutes to cool the beans. I've found you can drop in the next load pretty much immediately after the cooling is done, although you may need to flush a little heat out of the drum first.
* If you forget the timer, the roaster stops dead. This is no problem if you are there, since you can just restart it immediately, and it picks up where it left off. But if you have stepped away, the beans will scorch and smoke in 20 seconds or so. I now reset the timer before each roast as a matter of habit.

Roast Quality The roaster was originally spec'ed for 100 to 200 gram loads, now the literature has its top load at 300 grams. If you hold the drum temperatures below 260C (500F), which is the upper limit for lower density beans, a 300 gram load will take about 20 minutes to reach a rolling second crack. Commercial roaster manufacturers who specify roast times (Probat, Diedrichs, Ambex, Toper) spec their gas fired, ventilated drum roasters to do production roasts in 10 to 15 minutes, and table top electric models with lower ventilation at 15 to 20 minutes. I think 15 top 20 minutes is too slow for this roaster, since the airflow is higher than most electric models. So I added an insulating wrap to get roast times with 225 gram loads down to the 10 to 15 minute range. However, my ten minute cupping roasts have been slightly undeveloped; so it could be that the airflow on this roaster makes it better to target a 12 to 17 minute range for roasts. In that case, the drum insulation may not be needed if roast loads are kept below 225 grams.

Thanks!
It is nice that the cleaning and maintenance are so user friendly!