Civilizing the i-Roast2 (Part 1)

[hazbean]

Re: Civilizing the i-Roast2 (Part 1)


So with very similar setups, we both are ending up with close correlation between
programmed and measured times.

You do it with a resistor value of 2.7K at 246VAC line voltage.
I do it with a resistor value of 1.8K at 242VAC line voltage.

Interesting to contemplate the relationship between optimal R and VAC.
Not linear I suspect.

Did a roast at 10C ambient earlier. Still behaving well. Rained so #^%$# hard
that I could hardly pick second crack.

Have you done the Red Sea blend yet?

[ramking]

Re: Civilizing the i-Roast2 (Part 1)

Have you done the Red Sea blend yet?

 Not yet,Im working my way through a few other roasts.
Ill try a couple of Red Sea roasts at the weekend and let
you know how successful they were.
 I agree about the relationship of VAC to R.Im at work now
so will check the readings I took when I tried the same bean
roasts with diff R values.I dont think it was linear either but
will post again later.

Ken.

[justintime]

Re: Civilizing the i-Roast2 (Part 1)

Its doubtful you guys will see this post, but I was wondering what tolerance resistors you guys were using? Using a 10% resistor could result in values 200 ohms off. Ie 2k could really be 2.2k. Of course if you are measuring them that would not explain the difference in resistor values.

[Dimal]

Re: Civilizing the i-Roast2 (Part 1)

Its doubtful you guys will see this post, but I was wondering what tolerance resistors you guys were using?  Using a 10% resistor could result in values 200 ohms off.  Ie 2k could really be 2.2k. Of course if you are measuring them that would not explain the difference in resistor values.

Even if you opt for a 1-2% tolerance resistor, its still going to be awfully cheap mate. Just go to your nearest electronics supplies store and buy a few within a nominal range. Bound to get one that works almost perfectly for your machine then...

Mal.

[hazbean]

Re: Civilizing the i-Roast2 (Part 1)

As Mal suggests, I bought a few each of several values, and measured them with my
multimeter. DSE has them all in labelled bins. Except dont believe
the labels -- one of them was 10x out.

At one point I tried to relate the difference in reported values to
line voltage, but that doesnt seem to explain it either.

[chips60]

Re: Civilizing the i-Roast2 (Part 1)

Brought a 2nd hand IRoast (1 not 2) from Mark the Coffee Barun a few weeks back and discovered it runs hotter than expected. This is a cold Adelaide winter so I would hate to see the effect of high summer temperatures on the IRoast.

Pulled the cover off the base and swapped the 1k for a 2k2 resistor this morning whacked in 150gms of Indian Monsoon Robusta as a test on profile 2 and it now works a treat. Ran the full profile time without burning the beans (before I would of had smoke poring off the beans if left for the full profile, had to stop each roast at least 2 minutes early). The fan is now cycling which I never heard it do before. Very happy camper now. I have run 2 wires from the new resistor out beside the power cable to allow for further calibration with a pot if needed.

Will fit a thermocouple this afternoon and run roast monitor to record my next roast.

Many thanks to Hazbean and others for these posts.

Cheers
Greg.P

[justintime]

Re: Civilizing the i-Roast2 (Part 1)

Its a crime that they have a calibration resistor and never calibrate it at the factory. Whats the point? Someone needs to find a tried and true calibration method so we all can calibrate them the same. Im using a 3.6k but never checked it with a thermocouple. I just ran the first profile and it was a light roast. I could probably come down a bit. I dont know what would be better, that the screen matches the TC or the program temp matches the TC or the Screen?

Has anyone ran the machine without beans for calibration? Would it matter?
I hate to waist beans.

Larry

[hazbean]

Re: Civilizing the i-Roast2 (Part 1)

Greg, thanks for the kind words. Maybe we can catch up at one of
Marks 2pm Sat cuppings and swap notes, I try to get there as
often as I can. 2k2 is consistent with what I used. Well worth
using the TC; it is easily inserted from the bottom, and kept in
the same position each time (about 2cm above the bottom of
the chamber and half way across I thought was a good position).

Larry, I think my comments in another thread are pretty close
to yours! As for running the roaster without beans -- about
halfway down the first post in this thread, I mentioned that I
did that at a given resistor value in order to get an idea of heat
generation at that value. This gives an indication of behaviour
that can be confirmed with real beans (as in post 2).

I managed to match all three readings (TC, screen, program) at
later stages of a roast (not possible early on). Thats what I would
define as a "calibrated" iR2.

[justintime]

Re: Civilizing the i-Roast2 (Part 1)

Yes 2nd post somehow I glassed over that sentence. I guess I will try a TC this weekend.
thanks

[chips60]

Re: Civilizing the i-Roast2 (Part 1)

I saw your post about mounting the TC last week and found it dead easy to do. The DSE TC you suggested fits easily, the Jaycar one would of been a tight fit.

Roasted 150gm of PNG Sigri AA and here is the roast monitor data graphed. This was on Preset 2 and it ran for the complete 11 minute cycle.



Here is the post roast photo. Dont ask for the 1st - 2nd crack data as with my hearing and the roar from the IRoast I cant hear it.



Looking forward to drinking this one.

Time to start with some custom profiles me thinks.

Will have to get to a cupping session one saturday. I usually go to Marks shop during the week as I work at Regency Park. Live down south and dont get north much on weekends.

Cheers
Greg.P

[hazbean]

Re: Civilizing the i-Roast2 (Part 1)

A reasonable looking profile, I think. Nice to have the plotting
software going.

I found it hard to tell bean colour from the exposure, so I took
the liberty of tweaking your pic a bit (see below, hope its close).
I think they look good. No tipping I can see, a few blotches but
that I believe is not a problem. Will be interesting to find out
how it cups, may possibly even be a touch light. That would
be consistent with what I got for preset1 in the second post
of this thread. Will be interesting to see what happens with
other profiles.

Personally I think an iRoast delivering a proflle like this is a
pretty good roaster, just a pity its so fiddly to get it.

[Dimal]

Re: Civilizing the i-Roast2 (Part 1)

That ones a bit hard for me to see properly too "hb".... I know photos on uncalibrated monitors results in a very subjective appraisal of the finished roast but I can see the beans better with my "Gimp" enhancement : and like you say, generally it(the roast) looks pretty good.

Naturally, the only real test is going to come when a brew or three is made from them a few days down the track. Looking at your roast chart "chips60", the I-Roast certainly gets going in a hurry... kind of like a demented popper . Is there any way to slow that ramp down a bit more by any chance, just looks a bit too steep to me and getting to First Crack territory in about 4 minutes is very fast.

I know absolutely nothing about using the I-Roast so dont know what is possible but Im sure you will notice an improvement in the cup if you can slow down that initial ramp somewhat. All the best mate,

Mal.

[sdg]

Re: Civilizing the i-Roast2 (Part 1)

Naturally, the only real test is going to come when a brew or three is made from them a few days down the track. Looking at your roast chart "chips60", the I-Roast certainly gets going in a hurry... kind of like a demented popper . Is there any way to slow that ramp down a bit more by any chance, just looks a bit too steep to me and getting to First Crack territory in about 4 minutes is very fast.

I know absolutely nothing about using the I-Roast so dont know what is possible but Im sure you will notice an improvement in the cup if you can slow down that initial ramp somewhat. All the best mate,

Mal.

A "demented popper" -- love it 8-)

The iR2 seems specifically "designed" to go berzerk for the 1st few minutes -- ignoring all instructions and entreaties to the contrary. If Hearthware ever heard of the "slow ramp to FC" approach to roasting then they sure arent letting on. Leaving the chaff-collector off helps a bit, but not enough, and isnt very "adjustable" besides (its either on or off); and trying to slow down the so-called warmup phase (ha ha) with a beefier cal resistor I suspect would kill the rest of the roast (hazbean would know much more about this than I)

The only mod that really sounded plausible for this purpose (to me) is a pot with a nice dial across where the cal resistor is supposed to be -- then you get to sit there twiddling the dial to control the roast -- just like with an adjustable heat gun now I think about it.

At this point you (the demented iroast owner that is) may as well strip out all the crap that supposedly makes the thing programmable and post it back to Heathware with a nasty note, coz you no longer have a programmable roaster >

Unless you wanted to build your own controller, and do it properly...

[sdg]

Re: Civilizing the i-Roast2 (Part 1)

PS. My that roast looks good now the gimp has had its evil way with it. Who needs a thermocouple 8-)

[justintime]

Re: Civilizing the i-Roast2 (Part 1)

I did another roast with the 2nd profile expecting it to be a dark roast. After seeing hazbeans My 2nd roast is on the left. In other words it looks almost exactly like the 1st profile. Since my resistor is 3.6k I will lower it for the next test maybe 3.3k and see if it gets darker I already tried 2.6k which was too hot.  It appears that once you get over the hump (too high K) with the resistor it doesnt matter the value any more. I think what were trying to find is the exact center of the hump, where the increased value doesnt make a difference anymore.
I think starting with a high value and creeping down until the roast on profile 2 is the right color is the correct value. I will do another roast later today.
Here are pictures, roast on right is profile 1 and on left is 2. If anything 2 looks a tad lighter than 1

PS: it seems to help the color balance of the picture if you have something white in the picture for the cameras color balance to cue on..