| |
Main
Date: 22 Jun 2006 22:18:38
From: diab0lus
Subject: i-roast 2 profile adjustments for ambient temperature changes
|
I have been collecting temperature data for my roasts with a
temperature probe inserted through a small hole drilled in the top of
my i-roast 2. I have figured out a formula to predict, within a
certain degree of accurracy, the difference between the average
measured in-the-bean temperature (recorded every 15 seconds) of the
i-roast 2 and the average programmed temperature (every 15 seconds) as
they relate to the ambient temperature.
The temperature ranges that were used while collecting this data range
from between 50 F and 85 F. The formula predicts that the average
programmed temperature equals the average measured temperature at about
30 F. I recorded the measured temperature every 15 seconds throughout
the duration of each roast. All of my profiles are of the ramp-up
variety. I don't drop the temperature at any stage in my profiles and
don't know if this will work with those types of profiles.
Here is the formula to calculate the amount that the average measured
temp should exceed average programmed temp (F) as a percentage:
(ambient temp - 30)/1100
So if it is 90 F today, the % over average programmed temp will be
about 5.5%. You can use these figures to adjust your roast profiles
accordingly.
If you want to input your average programmed temp into the formula to
find out your average roast temp use:
([(ambient temp - 30)/1100]+1) * Average programmed temp = Average
roast temp
------------------
On another note, I have observed faily consistently that the i-roast 2
will exceed its programmed temperature by about 19%. So I have been
programming the last stage on the i-roast by using this formula:
Desired Roast Temperature / 1.19 = Programmed Temperature
Using this approach requires the stage to be at least a minute long.
It shouldn't be preceded by a big jump in temperature - it needs time
to biuld up heat. What I have observed is the approximate maximum that
it will exceed the programmed temp.
Using these two approaches I have been able to get about a 15 minute
roast at the temperature I want with a nice ramp-up temperature curve
in some different ambient conditions. If my wife let me roast in the
house, I wouldn't need to go through all of this trouble.
I am interested to hear if these formulae work for anyone else using an
i-roast 2. I have read that performance is inconsistent from unit to
unit.
Ryan
|
|
| |
Date: 23 Jun 2006 05:30:35
From: Hugh Jass
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
#@*& women!
I swear, if it wasn't for that little triangular patch of fur, there'd be a
bounty on 'em.
"diab0lus" <r0cketscientist@hotmail.com > wrote in message
news:1151039918.564613.175580@r2g2000cwb.googlegroups.com...
> If my wife let me roast in the
> house, I wouldn't need to go through all of this trouble.
>
> Ryan
>
|
| |
Date: 23 Jun 2006 17:46:16
From: diab0lus
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
I could brew a ground up brown paper bag and acheive the same results.
There are some things I wouldn't even wish upon my enemies.
I- >Ian wrote:
> <SNIP>
> > If my wife let me roast in the
> > house, I wouldn't need to go through all of this trouble.
> >
>
> Make her drink Folgers...
|
| | |
Date: 23 Jun 2006 23:02:53
From: Randy G.
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
LOL!!!
"Is this cinnamon, Dear?"
"No. It was a bag from Ralph's Market and they print their name on the
bag in that color."
Ground paper bag... that one's ging to come in handy! THANKS!!
Randy "see.. we can still laugh" G.
http://www.EspressoMyEspresso.com
"diab0lus" <r0cketscientist@hotmail.com > wrote:
>
>I could brew a ground up brown paper bag and acheive the same results.
>
>There are some things I wouldn't even wish upon my enemies.
>
>I->Ian wrote:
>> <SNIP>
>> > If my wife let me roast in the
>> > house, I wouldn't need to go through all of this trouble.
>> >
>>
>> Make her drink Folgers...
|
| |
Date: 23 Jun 2006 17:46:22
From: I->Ian
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
<SNIP >
> If my wife let me roast in the
> house, I wouldn't need to go through all of this trouble.
>
Make her drink Folgers...
|
| |
Date: 24 Jun 2006 20:23:38
From: Randy G.
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
"diab0lus" <r0cketscientist@hotmail.com > wrote:
I have read your finding a number of times, and it just does not
compute with me.. In one test I did, with a programmed roast, even
though the ambient temp was different, the measured bean temps were
almost exactly the same. I am not saying that your findings are
wrong- not at all. It is just that your findings, as I understand
them, seem to conflict with what I saw, and thus my confusion.
For my evidence of how my iRoast2 behaved, in my review:
[located here http://home.surewest.net/frcn/Coffee/iRoast2.html]
I posted the following graph:
http://home.surewest.net/frcn/Coffee/iRoast2/repeat.gif
It seems to indicate that even in varying ambient air temps the
machine was able to duplicate the roast profile I created, even
though, as we all know, the actual bean temps vary widely from
programmed temps.
>I have figured out a formula to predict, within a
>certain degree of accurracy, the difference between the average
>measured in-the-bean temperature (recorded every 15 seconds) of the
>i-roast 2 and the average programmed temperature (every 15 seconds) as
>they relate to the ambient temperature.
>
>The temperature ranges that were used while collecting this data range
>from between 50 F and 85 F. The formula predicts that the average
>programmed temperature equals the average measured temperature at about
>30 F. I recorded the measured temperature every 15 seconds throughout
>the duration of each roast. All of my profiles are of the ramp-up
>variety. I don't drop the temperature at any stage in my profiles and
>don't know if this will work with those types of profiles.
>
>Here is the formula to calculate the amount that the average measured
>temp should exceed average programmed temp (F) as a percentage:
>
>(ambient temp - 30)/1100
>
Here is my confused part- are you stating that at any point in a
program your formula applies, or...? I am confused because the
indicated temp, the measured bean temp, and the programmed temps vary
so differently at different times of the roast that any one formula
for predicting bean temp would seem to be futile as seen in this graph
from my review:
file:///E:/HTML/Coffee/iRoast2/Prog7graph.gif
I did drop the programmed temp some at the end, so what you stated
about that not working apply for that part of the roast, but I am
still confused at how the formula applies...
>So if it is 90 F today, the % over average programmed temp will be
>about 5.5%. You can use these figures to adjust your roast profiles
>accordingly.
>
>If you want to input your average programmed temp into the formula to
>find out your average roast temp use:
>
>([(ambient temp - 30)/1100]+1) * Average programmed temp = Average
>roast temp
>
maybe its the "Average" thing I don't get...? Average for any 15
second measured period? For the entire roast cycle duration of a
single programmed temp?
>I am interested to hear if these formulae work for anyone else using an
>i-roast 2. I have read that performance is inconsistent from unit to
>unit.
I would be glad to help you with some data gathering- it's been over
100 the last two days- that's a real LOAD of ambient!) but need some
help applying what you are finding.
It sure would be great to find an approach that would help folks
share data, as maybe the variations in iRoasts are due to ambient
temps and not differences in machines as we have suspected...? I don'
think so, but let's find out.
How about giving us the exact profile you are using, and we can post
our data to see if we all match...
Randy "my dog drank my roasting homework" G.
http://www.EspressoMyEspresso.com
|
| |
Date: 24 Jun 2006 20:21:52
From: Mike Hartigan
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
I have often speculated that a fairly consistent roast could be had
regardless of ambient temperature (with some limits at the low end,
no doubt) by adjusting the profile. The main impact that ambient
temp has on the behavior of the roaster is that lower temps slow down
the rate at which programmed temperatues are reached (on the ramp-
up). I have made a crude attempt to compensate for ambient temp by
programming 4 profiles, each for a different range of ambient temps.
For higher ambient temps, I insert an extra stage in between each
stage set to some intermediate temp, effectively mimicing the effect
of a lower ambient temp. I've had remarkably good results using this
method. It's interesting to actually have a formula with which I can
program my profiles to remove the trial and error approach. Thank
you for your effort.
WRT to the inconsistencies between machines, I suspect that your 1.19
factor is where the differences will lie. I may try to 'calibrate'
mine using your method and see where it takes me.
FWIW, I, too, have settled on a strict ramp up for my profiles. I
find that it's easier to follow the roast that way since the behavior
of the beans is much more predictable.
In article <1151039918.564613.175580@r2g2000cwb.googlegroups.com >,
r0cketscientist@hotmail.com says...
> I have been collecting temperature data for my roasts with a
> temperature probe inserted through a small hole drilled in the top of
> my i-roast 2. I have figured out a formula to predict, within a
> certain degree of accurracy, the difference between the average
> measured in-the-bean temperature (recorded every 15 seconds) of the
> i-roast 2 and the average programmed temperature (every 15 seconds) as
> they relate to the ambient temperature.
>
> The temperature ranges that were used while collecting this data range
> from between 50 F and 85 F. The formula predicts that the average
> programmed temperature equals the average measured temperature at about
> 30 F. I recorded the measured temperature every 15 seconds throughout
> the duration of each roast. All of my profiles are of the ramp-up
> variety. I don't drop the temperature at any stage in my profiles and
> don't know if this will work with those types of profiles.
>
> Here is the formula to calculate the amount that the average measured
> temp should exceed average programmed temp (F) as a percentage:
>
> (ambient temp - 30)/1100
>
> So if it is 90 F today, the % over average programmed temp will be
> about 5.5%. You can use these figures to adjust your roast profiles
> accordingly.
>
> If you want to input your average programmed temp into the formula to
> find out your average roast temp use:
>
> ([(ambient temp - 30)/1100]+1) * Average programmed temp = Average
> roast temp
>
> ------------------
> On another note, I have observed faily consistently that the i-roast 2
> will exceed its programmed temperature by about 19%. So I have been
> programming the last stage on the i-roast by using this formula:
>
> Desired Roast Temperature / 1.19 = Programmed Temperature
>
> Using this approach requires the stage to be at least a minute long.
> It shouldn't be preceded by a big jump in temperature - it needs time
> to biuld up heat. What I have observed is the approximate maximum that
> it will exceed the programmed temp.
>
> Using these two approaches I have been able to get about a 15 minute
> roast at the temperature I want with a nice ramp-up temperature curve
> in some different ambient conditions. If my wife let me roast in the
> house, I wouldn't need to go through all of this trouble.
>
> I am interested to hear if these formulae work for anyone else using an
> i-roast 2. I have read that performance is inconsistent from unit to
> unit.
>
> Ryan
>
>
|
| |
Date: 24 Jun 2006 19:49:48
From: Bob Wilson
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
diab0lus <r0cketscientist@hotmail.com > wrote:
THANK YOU!!
> I have been collecting temperature data for my roasts with a
> temperature probe inserted through a small hole drilled in the top of
> my i-roast 2. I have figured out a formula to predict, within a
> certain degree of accurracy, the difference between the average
> measured in-the-bean temperature (recorded every 15 seconds) of the
> i-roast 2 and the average programmed temperature (every 15 seconds) as
> they relate to the ambient temperature.
>
> The temperature ranges that were used while collecting this data range
> from between 50 F and 85 F. The formula predicts that the average
> programmed temperature equals the average measured temperature at about
> 30 F. I recorded the measured temperature every 15 seconds throughout
> the duration of each roast. All of my profiles are of the ramp-up
> variety. I don't drop the temperature at any stage in my profiles and
> don't know if this will work with those types of profiles.
>
> Here is the formula to calculate the amount that the average measured
> temp should exceed average programmed temp (F) as a percentage:
>
> (ambient temp - 30)/1100
>
> So if it is 90 F today, the % over average programmed temp will be
> about 5.5%. You can use these figures to adjust your roast profiles
> accordingly.
>
> If you want to input your average programmed temp into the formula to
> find out your average roast temp use:
>
> ([(ambient temp - 30)/1100]+1) * Average programmed temp = Average
> roast temp
I have a first generation iRoast and had noticed there was no
compensation for ambient temperature. But I didn't figure out the
correction curve. This is great!
I'll put together an excel spreadsheet for 5 degree increments from
40-90 and finally get back into the sweet spot, consistently.
> ------------------
> On another note, I have observed faily consistently that the i-roast 2
> will exceed its programmed temperature by about 19%. So I have been
> programming the last stage on the i-roast by using this formula:
>
> Desired Roast Temperature / 1.19 = Programmed Temperature
>
> Using this approach requires the stage to be at least a minute long.
> It shouldn't be preceded by a big jump in temperature - it needs time
> to biuld up heat. What I have observed is the approximate maximum that
> it will exceed the programmed temp.
>
> Using these two approaches I have been able to get about a 15 minute
> roast at the temperature I want with a nice ramp-up temperature curve
> in some different ambient conditions. If my wife let me roast in the
> house, I wouldn't need to go through all of this trouble.
>
> I am interested to hear if these formulae work for anyone else using an
> i-roast 2. I have read that performance is inconsistent from unit to
> unit.
I'll let you know in a month or so. I typically roast once per week and
it will take a couple of samples to tell. I had been stopped by 'heat
shrink' on my temperature probe but I've got a replacement that is
stainless steel enclosed.
Bob Wilson
|
| |
Date: 25 Jun 2006 21:21:38
From: diab0lus
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
> > From what I understand,
> >Hearthware's software takes over your profile until about 400F, then
> >the programmed profile will take over.
> >
> That is very interesting and something else I will have to look into.
My source was a thread on the Sweet Marias home roasting forum. You
are probably already aware of the searchable archive.
> I have found pretty much the same thing. Slower start and a total
> roast time of about 13-15 mintes or so makes for a smooth,
> well-balanced cup. other variables apply, of course. But the iRoast2
> was the first air roasted I have used that made a really nice roast
> for espresso.
Another recent topic on the SM forum (SM, not S&M) discussed that for a
given roast temperature, the length of the roast is inversely
proportional to the amount of acidity tasted in the cup for a given
bean/blend. Espresso drinkers especially supported the longer roast
times. Since i-roast 2 users are limited to 15 minute roasts, it makes
sense to me to develop profiles use all of that time.
If you use the 19% calculation, you can figure out the programmed
temperature for the end stage for the roast temperature that you want.
I round up for good measure. I don't want to hit 15 minutes with the
temp too low. 450F roast is approximately a programmed 375-380 on my
roaster. It seems that the extreme end of programmed roast
temperatures is where the real control lies for the IR2.
My next profile will involve smoothing out the jump from stage I 320 F
to stage II 370 F with a more linear increase. I haven't messed with
this end of the profile too much.
Another odd thing, I did two batches of Monkey, 150 g each. Same
profile, same ambient temp (one direct sunlight, one with cloud cover)
within 20 minutes of eachother. The temperature in the roast chamber
in direct sunlight rose to about 99 degrees before the roast started.
This roast completed a minute earlier than the first roast (both
roasted to 455 F). I think a warm roaster and direct sunlight
contributed to it finishing faster. I vacuum sealed both batches in
different bags. The roast that went to 13:30 gassed a little bit - the
beans were loose. The roast that went to 12:30 gassed the bag up
really tight, like a balloon. I wonder if the amount of CO2 is any
indication of the level of acidity.
> Right now the daytime highs have been n the 100-104 range, so I will
> try to do a roast tomorrow in those temps, then, when global warming
> settles down a bit, I'll do a backup roast. I have a dual logging
> digital thermometer, so getting thedata should be quite simple..
I will await your response patiently.
> It all seems like it should be so eays, no? AArghhh...
Hah. However, I am maker better and better roasts with this thing, so
something is paying off. I think it is the longer roasts with the
lower temperature.
I'm waiting for someone to chime in to say that I shouldn't be roasting
to temperature alone. When the beans smell yummy should I hit the
"cool" button regardless of temperature?
Ryan
|
| | |
Date: 25 Jun 2006 22:14:00
From: Randy G.
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
"diab0lus" <r0cketscientist@hotmail.com > wrote:
>> It all seems like it should be so eays, no? AArghhh...
^easy...
>
>Hah. However, I am maker better and better roasts with this thing, so
>something is paying off. I think it is the longer roasts with the
>lower temperature.
>
That is something I learned when I went from the Hearthware Precision
(which had as it's only control a mechanical timer and a stop/cool
button) to the Hottop. The longer, "slower" roast was far more
satisfyi9ng, muting the acidity and giving a more balanced cup. That
is a subjective thing, and depending on the coffee and the palate, as
well as the brewing method, "to each his own" is rarely been used more
appropriately. The Hottop's roast worked so well for me that I have
not roasted in an air roaster for some three or four years other than
the testing done for my website, and soon for this experiment of which
we speak.
As some here have tested with some sophisticated air roasters (like
Bob Yellin has done) that it is more a function of the roast time and
profile than it is of the roasting appliance. Additionally, Jeff
Pawlin with his computer controlled Hottop roaster and custom software
has demonstrated that even small changes in the roasting profile have
an impact on coffee quality. Few have experimented to that depth of
roast control because equipment that allows such minute control of the
roast is/was fairly well unavailable to us unwashed masses.
Still, with that in mind, the iRoast2 is a very capable home roasting
appliance for the money, and well capable of creating some really
great roasts, with repeatable results (to be proven), and roasts that
are quite comparable to the Hottop. And it points out the one lacking
feature (IMO) the Hottop, but that has been made more than clear by e
for years- here, on my website, and in person to the Hottop folks.
>I'm waiting for someone to chime in to say that I shouldn't be roasting
>to temperature alone. When the beans smell yummy should I hit the
>"cool" button regardless of temperature?
>
I would be one, having done a lot of home roasting to sound alone. On
the other hand, it is more difficult to do so with the iRoast because
of the Boeing roar it puts out, so temperature is a good indicator of
roast level as my testing showed, but it would take a fast-responding
probe located IN the bean mass that responds fast enough and is placed
to read the beans more and the incoming air less for accuracy. It
would also have to be mounted in a way that would place it in the same
location every time for repeatability.
>Ryan
What? err... ...No, I'm Randy... Oh... YOU'RE RYAN! ;-) Never mind.
Randy
"think sadly of all the people
missing this becasue of their killfiles...
Condescending... Moi?"
G.
http://www.EspressoMyEspresso.com
|
| |
Date: 25 Jun 2006 19:22:53
From: diab0lus
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
> I have read your finding a number of times, and it just does not
> compute with me.. In one test I did, with a programmed roast, even
> though the ambient temp was different, the measured bean temps were
> almost exactly the same. I am not saying that your findings are
> wrong- not at all. It is just that your findings, as I understand
> them, seem to conflict with what I saw, and thus my confusion.
How different was the ambient temperature between the two roasts? An
increase in temperature from 70 to 75 degrees only needs an increase of
0.45%. Probably not worth doing. However, if the ambient temperature
went from 45 to 85 between roasts, you would have a calculated overall
increase of 3.63%, which could be the difference of 14 F (an arbitrary
number based on one of my own profile temperature averages) at the same
point in the roast profile.
> For my evidence of how my iRoast2 behaved, in my review:
> [located here http://home.surewest.net/frcn/Coffee/iRoast2.html]
> I posted the following graph:
> http://home.surewest.net/frcn/Coffee/iRoast2/repeat.gif
> It seems to indicate that even in varying ambient air temps the
> machine was able to duplicate the roast profile I created, even
> though, as we all know, the actual bean temps vary widely from
> programmed temps.
This review helped me decide to buy the i-roast 2. What I don't know
is how varying the different air temperature are that you mention.
> maybe its the "Average" thing I don't get...? Average for any 15
> second measured period? For the entire roast cycle duration of a
> single programmed temp?
You should start with data from a roast that you are happy with in
whatever ambient conditions you were subjecting the roaster to at the
time. Record your measured roast temperature every 15 seconds from
beginning to end. Average all of the measured temperatures for that
roast. Note this temperature as target average roast temp. It will
probably be a number in the mid to high 300 F range.
Take your programmed roast profile, chop it into 15 second intervals,
average this number. If you have a 15 minute profile and you stopped
it at 12:30 during the roast above, only use the data through 12:30 to
create your average.
Note the ambient temperature. Run the ambient temp and average
programmed temp through the equation:
([(ambient temp - 30)/1100]+1) * Average programmed temp
This will give you an average roast temp number. Compare this number
with the target average roast temp noted above. If it is different,
modify your roast profile (in excel or whatever you use) by
increasing/decreasing the heat. Run the new average programmed temp
through the equation. When the numbers are close, your profile is
ready.
Note: Modifying the numbers in the beginning of you roast profile will
probably have little effect on the performance and may only serve to
provide unusable numbers from my formula. From what I understand,
Hearthware's software takes over your profile until about 400F, then
the programmed profile will take over. I have the first 5 minutes of
my profile set to 320F. (I started working with your profile 10 from
the above referenced review on your web site.) Any tweaks done to the
profile should be done after the measured 400F point.
> I would be glad to help you with some data gathering- it's been over
> 100 the last two days- that's a real LOAD of ambient!) but need some
> help applying what you are finding.
Hopefully I have clarified it a little. It makes sense in my head, I
just need to get it out - past all of the cobwebs.
> How about giving us the exact profile you are using, and we can post
> our data to see if we all match...
Today I used my profile 6 to reach 450 F in 11:30. The ambient temp
was 80 degrees (IIRC, it was a non-recorded roast).
0:00 - 5:00 320
5:00 - 10:00 370
10:00 - 11:45 375
11:45 - 13:30 380
13:30 - 15:00 385
(375 * 1.19 is 446, which is close to the temp I acheived)
I am discovering more and more good things about lighter roasts for
espresso blends, I will be adjusting this profile to end at 14 - 15
minutes at 450 or so.
Anyhow, I am interested in everyones feedback. I consider this to be a
somewhat perturbative approach to the problem, so I think I am on the
right track - I just might need to tweak it a little bit. If anyone is
getting nowhere with this formula, perhaps we can help determine if my
method is wrong or if there are differences between the i-roast 2s or a
combination of both.
Ryan
|
| | |
Date: 25 Jun 2006 20:08:35
From: Randy G.
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
"diab0lus" <r0cketscientist@hotmail.com > wrote:
>
> From what I understand,
>Hearthware's software takes over your profile until about 400F, then
>the programmed profile will take over.
>
That is very interesting and something else I will have to look into.
>>[Excellent and detailed reply snipped]<<
>I am discovering more and more good things about lighter roasts for
>espresso blends, I will be adjusting this profile to end at 14 - 15
>minutes at 450 or so.
>
I have found pretty much the same thing. Slower start and a total
roast time of about 13-15 mintes or so makes for a smooth,
well-balanced cup. other variables apply, of course. But the iRoast2
was the first air roasted I have used that made a really nice roast
for espresso.
>Anyhow, I am interested in everyones feedback. I consider this to be a
>somewhat perturbative approach to the problem, so I think I am on the
>right track - I just might need to tweak it a little bit. If anyone is
>getting nowhere with this formula, perhaps we can help determine if my
>method is wrong or if there are differences between the i-roast 2s or a
>combination of both.
>
Right now the daytime highs have been n the 100-104 range, so I will
try to do a roast tomorrow in those temps, then, when global warming
settles down a bit, I'll do a backup roast. I have a dual logging
digital thermometer, so getting thedata should be quite simple..
It all seems like it should be so eays, no? AArghhh...
Randy "the roaster is taking over my roast" G.
http://www.EspressoMyEspresso.com
|
| |
Date: 28 Jun 2006 09:59:15
From: finding z0
Subject: Re: i-roast 2 profile adjustments for ambient temperature changes
|
Randy G. wrote:
> "diab0lus" <r0cketscientist@hotmail.com> wrote:
>
> >> It all seems like it should be so eays, no? AArghhh...
> ^easy...
> >
> >Hah. However, I am maker better and better roasts with this thing, so
> >something is paying off. I think it is the longer roasts with the
> >lower temperature.
I have exp'ted briefly with extended roast times on my I-Roast1 w/o
temp probe and was disappointed by the results. If hearing cracks is
difficult with this machine (I think I can hear them on shorter roasts)
then lenghtening the roast complety obliterates them or spreads them
out. I bought a temp probe but have never used it. Something about
drilling the hole, or that the tip might be the wrong material got in
the way. I too seek the body of the roast that seems to be lacking in a
quick roast profile. I found that some beans just have more body/ less
acidity even if i screw up a bit. I'm taking heart that longer, slower
roast times (with temp control) do lead to the desired roasts. Cheers
|
|
