Vicki, Ryan:
I was curious also about whether the iR2 would decrease the temperature 
after the override period so I entered Ryan's 15 minute profile and roasted 
some Columbian Huila San Jose de Isnos.  The results appear to be pretty 
much a disaster.  The roaster did not reduce the temperature after the 
default period.  Not only that, but it never increased the temperature when 
the 2nd stage program called for 330F, so basically it maintained 350 
indicated for 9:45, ending stage 2 with a thermocouple reading of 394F. 
Stage 3 maintained an indicated temperature of 367, taking the beans to 410F 
at the end of the 4:15, the final stage maintained an indicated temperature 
of 376F, and took the beans up to only 416F.
My thermocouple reads 435F for a C+ (or so) roast, so this came out 
extremely light, uneven and only about half the beans appear to have 
completed 1st crack.  I will be interested to see what they taste like. I'm 
thinking it's gonna be baked, big time.
So it appears that as reported, the iRoast machines are very inconsistent, 
one from another, and in my case at least, no, the machine does not drop the 
temperature down to the programmed temp if it is lower than the default 
stage.
It also appears that the machine either:
1. ignores program increases of less than some amount, or
2. has some unknown break points that your program temp has to fall between, 
or
3. has some unknown break points that it can control to.
Indicated (LCD readout) and ThermoCouple readings at 30 sec intervals:
Time    Ind    TC
Prog 320F
0:30    291    175
1:00    325    245
1:30    342    289
2:00    352    314
2:30    350    331
3:00    350    345
3:30    350    355
4:00    352    361
4:30    350    366
5:00    350    372
5:30    352    376
6:00    352    381
6:30    350    383
7:00    350    388
Prog 330F
7:30    350    389
8:00    350    389
8:30    352    391
9:00    356    391
9:30    350    391
9:45    352    394
Prog 350F
10:00    367    396
10:30    367    400
11:00    365    403
11:30    367    405
12:00    367    406
12:30    365    408
13:00    367    409
13:30    365    410
14:00    367    410
Prog 365F
14:30    376    414
15:00    376    416
This is not a big disappointment for me, I'm currently pretty happy with my 
8:30 to 12:00 roasts to C+ - FC or so.  I just wanted to know what my 
machine would do and find out how much variation there is between machines. 
Looks like there's a lot!
Michael Wade

Hi Michael,
I was pretty sure Ryan's profile wouldn't be a good one for my machine 
either. So far the ones that work best for my machine seem to start at 
350, increase (or decrease) by at least 20-25 degrees for each step, and 
have at least 2 minutes on any one step. A decrease seem to need at 
least 3 minutes to really change the way the roast develops.
I know they are all different, but these general guidelines seem to work 
for my IR2.
I'm also finding that roast size makes a huge difference. I would assume 
that smaller bean masses are more responsive to changes in temperature 
settings, with bigger batches holding on to heat and developing more 
heat regardless of any small changes I make.
vicki
Michael Wade wrote:
<Snip>

My two iRoasts (orig. and 2) are like night & day. The original one  
runs so hot that even with a long extension cord, it'll overheat and  
shut off if I use one of the presets--or even program a multi-stage  
roast too hot. The 2 runs very true to temp and has always done well  
with both the presets--only time it ever overheated and shut off was  
with that disastrous load of fermented JBM greens from a non-SM  
vendor (they were so heavy that they did not circulate well, allowing  
heat to build up in the bean mass to the shutoff point). And I have  
yet to stall a roast in it yet--I even nearly did that in the SC/TO  
with some Harar Horse Lot 30 when I turned the heat down to 350 too  
soon (but I added 5 min. at 450 and saved the day).
On Jul 2, 2006, at 6:53 PM, Michael Wade wrote:
<Snip>
Sandy Andina
www.sandyandina.com

Michael,
FWIW, your indicated temperatures for programmed temperatures of 350F  
and 365F match mine exactly. I do not have any records from my  
attempts to go below 350F. I have managed higher starting  
temperatures. A programmed temperature of 455F for the first five  
minutes resulted in an indicated temperature of 430F.
 From your results using Ryan's profile, and without any supporting  
measurements of my own, perhaps 350F is the minimum allowed  
temperature for the entire profile with some variation due to input  
voltage. My household voltage is typically 127 volts, which is a bit  
high. I would complain to my electric company, but that is who I work  
for.
I think someone on CoffeeGeek mentioned that the iRoast heating  
element only has three temperatures (low, medium, and high), with the  
fan providing the remaining temperature control. Assuming two or  
three possible fan speeds, that would give 6 or 9 achievable  
temperatures, including the minimum and the  maximum.
On Jul 2, 2006, at 5:53 PM, Michael Wade wrote:
<Snip>

Hi Vicki,
<Snip>
IMO, the machine is very temperature-agile on temperature increases.  If it 
is so inclined, it can raise the bean mass temperature pretty rapidly, but I 
am beginning to think of it as having the spirit of a cat.  It is touch and 
go whether it will feel like cooperating on any given day.  It was not the 
least bit interested in Ryan's profile.
The only time I've tried a decrease was when I was roasting a fast profile 
and programmed a finish stage of 355 that acted as a slow cooldown to make 
sure the beans were temperature soaked all the way through.  You're right, 
it took a while to get the bean mass temp down.
<Snip>
I always use a weighed batch of 130 grams, simply because that fits in my 
clamp-top glass canning jars and eliminates a variable.
I roasted several batches today besides the 15 minute test, and I (finally!) 
like what I'm seeing in terms having some control of roast depth, at least 
at C+.  I am impatient (never a productive learning atmosphere) to get 
enough confidence to try roasting some of the Triple-Pack that I bought. 
Maybe later this month.
The whole house smells like coffee.  I love it!
Michael

Sandy,
<Snip>
Oh, I remember that story.  I've always wondered what those beans were 
coated with or soaked in, and for what purpose.
Ever figure that out?
Michael

David,
<Snip>
Whaddya know, at least two of them are close!
<Snip>
Even for the first couple of minutes?  When I program something like 400F 
for the first 5 minutes, the machine won't go over 350 for some indefinite 
period.  I've seen it last as little as 1:50 or as long as 2:50.
<Snip>
That was an unexpected result, for sure.  Not only did the machine not track 
the programmed temp after the default was satisfied, but either the 10F 
increment was not sufficient to get the control program's attention or there 
is some predetermined temperature point that the processor is looking for 
before ratcheting up the controlled temperature.  Maybe 11F, or 13, or 15 
would have done the trick.   I'm not feeling ambitious enough to design a 
series of tests to find out. 
 I hadn't thought about it being related to line voltage.  I just went out 
and checke the appliance receptacle that I use for roasting;  it's 122.9V 
open circuit, and 121.4V loaded by a toaster.  seems OK.
But you would know better than I!
<Snip>
<Snip>
That's interesting.  I'll have to do an exploratory.  Did they say whether 
there are multiple elements that are switched in and out, or any guess as to 
how the heat is controlled?  I would think that anything other than 
switching would be prohibitively expensive.
Given the undocumented default override and the apparent inconsistency 
between machines I have to wonder what other undocumented changes have been 
incorporated between (or even during) production runs.  It woujld be 
interesting if we could somehow find the production date for our respective 
machines and correlate that with the other symptoms.
Then again, sufficient unto each day the evil thereof...  don't go looking 
for trouble.
Michael

This is the profile I find myself using most often:
350 3 minutes
375 2 minutes
425 3 minutes
450 3 minutes
I think one of the reasons it works so well for me is that the fan speed 
changes almost exactly (mebbe a 10-15 second lag)when the temperature 
changes kick in.
Yesterday I roasted 150 grams to City+ and stopped the roast with 2' 15" 
remaining. That seems pretty typical, regardless of the beans (within 
15-20 seconds). This profile gets me to Vienna (as far as I have ever 
gone) with between 45 seconds and one minute remaining, depending on the 
bean.
Bigger batches (150 grams) take longer to get to first crack, but the 
total time elapsed is about the same as roasts that start with 130 
grams. The difference between batch sizes is that with 130 grams the 
time between 1st and 2nd crack is longer.
What I would like to do next is setup a profile that increases the time 
at step 2 or step 3 see if I can prolong the time between first and 
second crack a bit. I somehow think that perhaps because of the size of 
the roast chamber vis-a-vis the bean mass, this is less doable than it 
might appear,
I'm very satisfied with the roasts I am getting, but I wonder if I 
couldn't get a little extra magic if I could do that.
Vicki
!*!*!*!*!*!*!*!**!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*!*
                 Taming Coffee: The Weblog
                 http://taming.motime.comMichael Wade wrote:
<Snip>

On Jul 3, 2006, at 1:45 AM, Michael Wade wrote:
<Snip>
The ramp took about 1 minute. The machine was at 430F at one minute  
and it settled to 433F at two minutes. I used this profile several  
times early on, but I only have one set of measurements.
<Snip>
Given what I have seen from using this machine, I think the limited- 
number-of-available-temperatures idea is the correct one. I have a  
lot of strange profiles that bear this out, but something else may  
happen that makes me change my mind. If this idea is correct, then  
the temperature you end up with depends on where the programmed  
temperature falls in relation to one of the available temperatures  
and the slop in the controller. On my machine, programmed  
temperatures of 350F and 360F both result in indicated temperatures  
of 367F, but a programmed temperature of 365F gives me an indicated  
temperature of 376F. So the trick to programming the iRoast is to  
find out what the available temperatures are and then set the program  
to hit those temperatures while not worrying about the differences  
between what is programmed and what is indicated.
Your voltage measurement is interesting. I was concerned that my 127  
volts was causing the machine to run hot. That seems not to be the  
case, and I can avoid buying a variac.

Vicki,
What indicated temperatures do you get with this profile, with the  
measurements taken at least every 30 seconds?
On Jul 3, 2006, at 6:40 AM, Vicki Smith wrote:
<Snip>

Sorry, David. I never took 30 second outlet temps, and now with my 
machine behaving oddly, I can't duplicate this any more. The shame of it 
all, is that if Hearthware decides to replace the machine, chances are 
it won't duplicate it either. The machines seem incredibly variable.
The minute temps were pretty much spot on, though stage 3 ended at 320, 
and I never roasted all the way to the end of stage 4.
Had I known my machine would go flaky, I would have taken more frequent 
measurements. I might have a relative lack of multi-tasking ability, but 
I found that when I tried to take more frequent temps, I couldn't 
concentrate on the other things I was trying to attend to--the whole 
sounds, sight, smell thing.
Vicki
David Schooley wrote:
<Snip>

On Jul 5, 2006, at 5:24 PM, Vicki Smith wrote:
<Snip>
The advantage of 30-second measurements is that you have some slack  
if your transitions between stages also occur on one-minute  
intervals. I have found that the machine takes a minute or so for the  
temperature to settle during the first stage. The transitions for the  
remaining stages take 15 seconds or less. If you avoid sampling the  
temperatures during a transition from one stage to the next, then you  
should be fine. I use 15-second intervals because I like making  
graphs. Typing the numbers into a spreadsheet gives me something to  
do while the beans are cooling.