topbanner.gif
Login
Username

Password




Not a member yet?
Click here to register.

Forgotten your password?
Request a new one here.
Shoutbox
You must login to post a message.

renatoa
08/10/2022 1:56 AM
nguyencoffeesupply
and RoyB, Welcome

renatoa
08/06/2022 1:31 PM
Welcome, oak202, CoffeeNutZ and Mlcharlestonsc

Strangeworth
08/04/2022 9:34 PM
Thank you!

renatoa
08/03/2022 2:09 AM
Strangeworth and BigPalm, welcome to forum

Tamesino
08/02/2022 12:01 PM
Thank you guys and gals !! Exited to be here greenman

In Memory Of Ginny
Donations

Latest Donations
dmccallum - 10.00
JackH - 25.00
snwcmpr - 10.00
Anonymous - 2.00
Anonymous - 5.00
Users Online
Guests Online: 8

Members Online: 0

Total Members: 7,627
Newest Member: nguyencoffeesupply

View Thread

Who is here? 1 guest(s)
 Print Thread
make a programmer for the gene cafe
pauvrepiki
Hello,

I have a "Gene cafe" to roast my coffee for personal consumption. I found it preferable to apply the profiles of cooking depending on the type of coffee. Follow profile is quite restrictive because it is necessary to monitor the timing and temperature for about twenty minutes. So I missed several roasts because I do not always have the patience to keep eye on my machine.:@
I would like to replace a part of the original electronic machine by a robot microcontroller in order to select a preset profile rather than just temperature and time parameters.
Example: a profile: Malabar: Profile 1, Yrgacheffe profile 2: etc. ....
One could even consider the possibility of programming levels of cooking and store them in a memory cell as some hours heating programmer
But before embarking on this project, I wonder if there are people who have tried to change their "gene cafe". if there are schemes .... In short, if you have tried this experiment, could you share your knowledge on this subject? Thank you in advanceGrin
 
Koffee Kosmo
Welcome mate

This would be all well and good & a great idea
However growing variables from year to year also bean age and moisture content of the beans wont follow your rules

Its the nature of the beast called coffee
It will through a curve-ball when one least expects it

KK
I home roast and I like it. Designer of the KKTO
Roaster Build information
https://homeroast...ad_id=1142

https://docs.goog...lide=id.i0
Blog - http://koffeekosm...gspot.com/

Bezzera Strega, Mazzer Robur Grinder, Pullman Tamper Convex,
(KKTO) Turbo Oven Home Roaster.
 
John Despres
Yep, KK has nailed it. No coffee is the same.

However, there aren't many folk who have automated their Gene Cafe - in fact, none that I know of.

The beauty of this roaster is it truly teaches you to roast coffee because you have to pay attention to every sound, smell, visual cues, and other nuances that produce wonderful cups of coffee.

The risk of fire is pretty good - one should never walk away from the roaster. It's a tough concentration game at times, but the coffee is worth it!

John
Respect the bean.
John Despres
Fresh Roast 8, Gene Cafe, JYTT 1k, Quest M3, Mazzer Mini, Technivorm, various size presses and many more brewers.
 
oldgearhead
Control of your roaster would be real easy with a PLC such as an A/B SLC 504 (easily found and reasonable price on the market because they use outdated networks). You could program ROR profiles, ramp/soak, or invent your own math scheme. However, I would start with trying to figure out just how to install a bean mass temperature probe in the Gene'...
oldgearhead attached the following image:
dsc_9153_1.jpg

Edited by oldgearhead on 10/08/2011 6:01 PM
No oil on my beans...
 
pauvrepiki
Thank you for your answers. I make my request in another way: "Is it possible to get the schematics of the gene cafeShock
 
John Despres
The off-axis rotation makes a bean probe pretty difficult if not impossible, I think.

I don't know of any place to get a set of schematics.

John
Respect the bean.
John Despres
Fresh Roast 8, Gene Cafe, JYTT 1k, Quest M3, Mazzer Mini, Technivorm, various size presses and many more brewers.
 
gene
John:

For years I have been crossing my fingers that someone of the ilk of Bill or some other electrical genius would be using a Gene Cafe. Then we could get some mods.
Looks like its not going to happen. Tim would definitely get you a schematic. He could possibly get some answers. If anyone hears from him please post:
Fresh Beans, Inc. ( http://www.freshbeansinc.com ) - U.S Warranty Repairs Tim Skaling Phone: 435-940-1616. Fax: 435-940-1964

Thanks,
gene
 
oldgearhead

Quote

gene wrote:
John:

For years I have been crossing my fingers that someone of the ilk of Bill or some other electrical genius would be using a Gene Cafe. Then we could get some mods.
Looks like its not going to happen. Tim would definitely get you a schematic. He could possibly get some answers. If anyone hears from him please post:
Fresh Beans, Inc. ( http://www.freshbeansinc.com ) - U.S Warranty Repairs Tim Skaling Phone: 435-940-1616. Fax: 435-940-1964

Thanks,
gene


gene, i don't think the main problem is electrical. The challenge is as John stated, the off-axis rotation makes it very difficult/impossible ti install a BMT probe.
No oil on my beans...
 
lmclaren
regarding the BMT probe, I dont know the internals of the Gene Cafe, but could you use something like this?
http://www.instru...hermometer

You could mount it so that it look in through the side at an angle?

You may have to put in some logic to only read when the paddles are out of the way but the thermometer has a extremely quick read time so that would not be hard.
 
oldgearhead
I have a couple of IR meters (raytec brand) and they will only read the glass surface, not through it..
No oil on my beans...
 
lmclaren
That's correct, normal glass is opaic to IR.

I was thinking that you would make a hole in an area that the beans don't hit and then aim the IRT at the beans. My be able to look in through the fill hole.

Some of the IRT have lenses that focus the point to less than 1.5mm before it spreads again. You can also get windows with air purge to look through, special (silica?) glass that is transparent to IR.
 
mikepetro
I just ordered my Gene Cafe, its due in by the end of next week. I did a lot of homework on it though and I am fairly certain that I can do something that might satisfy the original posters request for about $150 in parts. My theory is actually very similar to something a guy in Australia did (http://tinyurl.com/8xjel5v), though my version would be a little different.

Basically I will use a PID controller with a Ramp/Soak feature. The Ramp/Soak feature is critical as not all PID controllers have them. I will use the controller to control a SSR that I will use to power the heater instead of the GC Heater Relay circuit board, ie a simple switching of 2 wires. I will put a Thermocouple in the exit air port to measure the air temp. Would love to measure bean temp but dont see how with the GC design. But hey, if the GC can control by air temp, this controller will be able to do at least as good only with saved programmable steps.

The PID controller I selected is this one: http://tinyurl.com/7bukxxu
This controller is nice because it has the ramp feature and it has communication capability built into it that allows it to talk to a PC. They even offer free software to program the steps and to plot\save the actual measured temps measured over time.

Soooo, I should be able to save various profiles, with up to 60 steps each (I would assume that only maybe 5 steps are needed) on my laptop and then send them into the controller at will.

The time/temps I give are for illustration purposes only, they do NOT represent a real roast.

I envision something like this. For a given profile I might have one step for preheating the machine at 200f for 5 minutes then add the beans, the next step for drying the beans at 300f for 6 minutes. Step #3 might be the Pre-1st_crack step where the temp increases to 479f for 8 minutes OR until I hear 1st crack. The next step might be the Post_1st_crack_slowdown where the temp is dropped to 461f for 3 minutes, then the cooling step kicks in.

During the roast the PID controller sends all the data back to the laptop and plots it for historical purposes. You also can make any time/temp changes at any any point to any step through the laptop or the controller directly.

Now, you wont always know in advance exactly what time 1st or 2nd crack is going to occur on a new bean so you might have to manually overide the programmed time by holding or advancing certain steps by ear till you get a specific bean down pat, i.e. when you hear 1st crack hit one button to manually advance from step 3 to step 4, the exact time would be saved on your plot so you can adjust your next profile if desired. There are Hold/Advance buttons on the controller to accommodate this.

I would envision building several profiles to fit different bean catagories, like maybe one for dry processed beans with a longer drying step, or one for Columbian FC roasts, another for Yemen FC+, etc. Yes, as was said earlier in the thread, you would need to "tweak" these profiles to each new batch of beans, but the various profiles would give you a good REPEATABLE starting point based on your roast history logs. Plus, if you bought 10 pounds of a certain bean you could dial in the profile precisely the way you like it and then reproduce it pretty darn accurately as long as you keep your batch size the same.

The one thing I know that I dont know yet is where to tap into emergency shutdown signals for blocked airflow. I will need to analyze the machine to figure that out unless someone here can point me in the right direction. I assume there is a sensor in there somewhere for airflow as I have heard of the GC shutting down when things get clogged. I would think that once I find the signal I could use it to shutdown the PID controller's output. The PID Controller has inputs for emergency shutdown.

Does anybody see any holes in my theory?

I will take photos and post my results here if there is interest.
Edited by mikepetro on 02/25/2012 9:46 PM
Mike Petro
Martinsville, VA
--------------------------------
Lady Silvia, Stepless Rocky, Gene Cafe
http://mikepetro.org/coffee/
 
gene
Mike:
Been away for month or so-When I checked in tonite yours was first post. How Lucky.
Yesterday I ordered a Thermapen handmade in England. But apparently it has to be directly inserted in airstream to work........ .......http://www.thermoworks.com/blog/2010/12/thermapen-made/

I'm planning to use it for espresso machine and steaming, etc. Was also hoping to probe around on the GC to see if can get any accurate readings, perhaps insert it in 3" exhaust tube I've attached to chaff collector. I even use this mod outside in summer. Really helps with cracks.

If you could use Thermapen would be glad to loan it.

And let me know if you get the schematics. Check in often and let me know if I can help. Unfortunately I'm a lot more mechanical oriented rather than electrical.
gene
 
mikepetro
Hi Gene,

Thanks for the offer but I already have the larger handheld version of the Thermapen. I bought it years ago for measuring meat temp when I smoke BBQ. Basically the whole Thermapen lineup uses Type K thermocouples which are well suited to our needs. The neat thing about the handheld version is that you can plug in any Tpye K probe. These Type K thermocouples are pretty cheap too, so I buy whatever one suits my need. Unplug the meat probe and plug in a coffee probe. GREAT for testing! For example here is a pic of my Therma 3 handheld with a small probe plugged into it that I use on my homemade SCACE device for calibrating my espresso machine.
mikepetro attached the following image:
dsc02697.jpg

Mike Petro
Martinsville, VA
--------------------------------
Lady Silvia, Stepless Rocky, Gene Cafe
http://mikepetro.org/coffee/
 
allenb
mikepetro posted:

I envision something like this. For a given profile I might have one step for preheating the machine at 200f for 5 minutes then add the beans, the next step for drying the beans at 300f for 6 minutes. Step #3 might be the Pre-1st_crack step where the temp increases to 479f for 8 minutes OR until I hear 1st crack. The next step might be the Post_1st_crack_slowdown where the temp is dropped to 461f for 3 minutes, then the cooling step kicks in.

I think you've successfully described a control shortcoming that has bugged the hell out of many of us around the globe. Trying to control a roaster with one set of steps and PID settings is like trying to herd cats.

My dream control setup would incorporate what you've described above -- being able to push a step advance button at the correct time during the transition at the end of drying stage, at 1C etc to kick in the different RoR steps.

One other "nice to have" feature of the dream controller would be the ability for each step to have it's own set of PID settings. At least in my experience, settings for the main roast phase are incompatible with the post 1C phase and I get lots of over and undershoots due to slow response times.

It's good to hear you're zeroing in on improved control solutions for the Gene. BTW, control improvements for one roaster will most likely benefit many other roasters as well.

Allen
Edited by allenb on 02/26/2012 11:16 AM
1/2 lb and 1 lb drum, Siemens Sirocco fluidbed, presspot, chemex, cajun biggin brewer from the backwoods of Louisiana
 
mikepetro

Quote

One other "nice to have" feature of the dream controller would be the ability for each step to have it's own set of PID settings. At least in my experience, settings for the main roast phase are incompatible with the post 1C phase and I get lots of over and undershoots due to slow response times.


Hmmm, I will play with that. While the controller doesn't have that feature built into it, I do have an idea.....

Blue sky time:
What if you were to autotune the controller at each step and record the ideal PID settings for each step. It's only 3 values. While running a laptop for monitoring you can change the individual PID settings quickly on the fly. So as each new step kicks in just punch in the 3 values that work best with that step. Should be able to do that within a few seconds, theoretically quick enough to stabilize the hysteresis for that stage of the roast.

Here is a screen shot of everything that can be changed on the fly while your ramp is running:
mikepetro attached the following image:
capture.jpg

Edited by mikepetro on 02/26/2012 12:44 PM
Mike Petro
Martinsville, VA
--------------------------------
Lady Silvia, Stepless Rocky, Gene Cafe
http://mikepetro.org/coffee/
 
JimG

Quote

mikepetro wrote:

Quote

One other "nice to have" feature of the dream controller would be the ability for each step to have it's own set of PID settings. At least in my experience, settings for the main roast phase are incompatible with the post 1C phase and I get lots of over and undershoots due to slow response times.


Hmmm, I will play with that. While the controller doesn't have that feature built into it, I do have an idea.....


Are you sure it won't accept multiple PID values? I ask because that display looks like the software that comes with the Delta DTB controllers (also sold under some other labels). And I'm pretty sure I remember that the Delta's will allow you to program different PID values at different SV's.

Jim

EDIT - on the screen shot, there is a drop list in the "PID Parameter:" group. I think that drop list is where you pick the PID parameters group that goes with that SV.
Edited by JimG on 02/26/2012 5:04 PM
 
mikepetro

Quote

Are you sure it won't accept multiple PID values?


Hi Jim,

NO, I am not sure at all. You may well be right on target. My controller arrives tomorrow and the software wont let you do much until it establishs a comm link to a controller so I havent been able to explore all its features yet.

I saw the PID Parameter group field but the software manual doesnt document its use. If you look at the Ramp Editor, all it gives you is an entry for Set-point and Time, there is no field to associate a step with "Parameter Group" in the Ramp file. The controller manual gives a little info but its not conclusive, to me anyway, until I can play with it and confirm its behavior. "If" it does do what we hope, it would allow you to enter 4 different setpoints and the PID parameters that match each one. It looks like it "might" do that with ramps too, and automatically use the nearest group to your setpoint but it doesnt say so in plain English anywhere. It only references "Operator Input" which seems to imply a manual SP change.

I will experiment tomorrow though B)

Here is a link to the reference (page 3-7) to parameter groups in the controller manual:
http://www.automa...lm/ch3.pdf

Here is the link to the software manual:
http://www.automa...lm/ch6.pdf


-----------------------------------------------------------------------
Edit
Looked up your Delta DTB and it does appear to be the exact same controller with a slightly different faceplate, though the parameters and functionality appear to be identical. The manual for the DTB is even worse than the Automation Direct one though.... uggh....
mikepetro attached the following image:
capture_1.jpg

Edited by mikepetro on 02/26/2012 5:54 PM
Mike Petro
Martinsville, VA
--------------------------------
Lady Silvia, Stepless Rocky, Gene Cafe
http://mikepetro.org/coffee/
 
JimG
Page 3-7 says that you can set the controller (by using PID4 as your parameter group) to automagically select the parameter group that is closest to the current SV. I don't think I ever actually went to the trouble to tune mine across a range of setpoints. But I do recall thinking it was pretty cool that it had that feature!

Jim
 
Vidget
Hello
sorry for my english

best way to make GeneCafe profiled is replace gene electronics by TC4 all_in_1 version with Artisan software. Easy, cheap, effective.
 
mikepetro

Quote

Are you sure it won't accept multiple PID values?


Well, I have good news and bad news. I got my controller in today and set it up on the bench to test. You can indeed plug in 4 sets of PID parameters, and the PID4 does work to auto select the closest one to your Setpoint. But there is a catch..... It only works in PID mode and does NOT appear to work in Ramp/Soak mode. In Ramp/Soak mode it "appears" to maintain the last used set of PID Parameters. Or at least the fields get greyed out and the numbers dont change as the ramp changes the setpoints. I guess it could be that it does change but just doesnt change the display, I am going to ask about this on the tech support forum.

Now you can easily plug in 4 sets of PID parameters and manually change your setpoints based on time and/or audio/visual cues (like 1C) and it will work beautifully by changing the tuning to match each setpoint change, but all setpoint changes would have to occur manually. This is probably going to my first mode of operation.

One other little quirk I ran up on is that I could have sworn I read that you can manually advance a ramp step, but I have yet to figure out to do it. Another question for tech support.........
Mike Petro
Martinsville, VA
--------------------------------
Lady Silvia, Stepless Rocky, Gene Cafe
http://mikepetro.org/coffee/
 
mikepetro
Imagine something like this http://www.phasei...nsors.aspx where you just toss this miniature wireless RFID device in with the beans and log the data......

The top end of this particular device is only 180c, I found another one with a top end of 200C, I bet that someone makes a 300C version.

Stay tuned......
Edited by mikepetro on 02/29/2012 4:34 AM
Mike Petro
Martinsville, VA
--------------------------------
Lady Silvia, Stepless Rocky, Gene Cafe
http://mikepetro.org/coffee/
 
kernelhmm

Quote

mikepetro wrote:
One other little quirk I ran up on is that I could have sworn I read that you can manually advance a ramp step, but I have yet to figure out to do it. Another question for tech support.........


Hi Mike,

Just found this fourm and people discussed the PID mod of gene 3D. I already done several preparing works.

1. Installed an adittion temperature sensor to share the same hole with the original one. (To leave the sufficient space, I have to remove the rubber shell of the original one )

2. breaked the heater circuit and bring two wires out for the additonal dimmer & SSR control.

Now I plan the following modifications:
1. a 2 way swith to choose between dimmer(hand) control and PID control
2. a bypass switch to short-cut the additonal circuit, by which to return the control back to the gene-3D original circuit.

Unfortunately, I bought a wrong PID, a panasonic KT-4 without the soak/ramp feature. So I have to manually write my own software for the profiling(try to dynamically change the PV settting by computer). Do you know if there is any open-souced data logger could be used to reduce the programming efforts? Thx
Edited by kernelhmm on 03/06/2012 2:07 AM
 
mikepetro
Hi again folks,

I havent forgotten about you all, just been trying to learn the GC and its default capabilities before I modified it too much. Not to mention learning how to roast in it. So far every roast has been better than the last, with no "bad" roasts yet, so I am VERY happy with the machine.

As for automating profiles, here is where I landed to date.

First, I decided that I wanted my Temp Controller to be a multipurpose device (echos of Alton Brown rattling through my brain). So I built a controller in a portable enclosure that can be used for:
-Automating GC profiles
-A Sous Vide controller
-A Smoker draft controller
-Profiling and calibrating my espresso machine with a SCACE
-The evening Prime Rib or just about any other Temp Control need.

This is what I came up with:
www.mikepetro.org/coffee/images/GC/DSC02705.JPG
You can plug any Type-K thermocouple into it, either standard or mini connectors. There is a standard electrical outlet on the front that you plug your heater, fan, whatever into. There is a data port to allow configuration and datalogging. All in a rugged portable box.

On the GC I decided to avoid permanent changes until I learned the machine better. The only permanent change was drilling a hole for the thermocouple right next to the existing probe. I routed the existing power wires out to an external switch (see gray box in controller photo above) where I can switch from PID mode to original GC mode at will. I relocated the fuse holder to inside of the GC to allow a hole for all the wires. All of this can easily be undone if I choose. The only downsides is if the fuse ever blows I have to take the cover off to get to it, AND, there is no air interlock for PID mode. In other words if the switch is set on PID and the fan of the GC is not on I could have a melt down. I will be careful to avoid this and I can easily make the interlock once I decide to permanently modify the GC.

www.mikepetro.org/coffee/images/GC/DSC02708.JPG


Here is the finished setup ready for me to begin serious testing:
www.mikepetro.org/coffee/images/GC/DSC02700.JPG
(Note, I live alone so I didn't get any grief for commandeering the Dining Room table)

With this I can:
-Control either via GC or the PID
-Datalog all temps throughout the roast
-Make config changes through the laptop
-Program pre-defined Profiles


One thing I have discovered after datalogging just one roast is that my added thermocouple responds MUCH faster than the GC temp probe. With the GC controlling there was some oscillation around the setpoint which is probably due to the slow response of their probe, or maybe measured value smoothing in their software.

I haven't attempted control via PID yet but I am sure I will be able to achieve more accurate results, plus have the advantage of Profile Programming.

Stay tuned.......
Edited by mikepetro on 03/11/2012 7:14 PM
Mike Petro
Martinsville, VA
--------------------------------
Lady Silvia, Stepless Rocky, Gene Cafe
http://mikepetro.org/coffee/
 
mikepetro
Oh my GOSH.... The results were even more dramatic than I anticipated. It appears the GC has a full 10 degree deviation above and below the setpoint. Thats a full 20 degree swing! Once I switched to PID it was LESS than a half degree deviation, less than a half degree total degree swing even!

Proof is in the data. I set the GC up to run empty at 400 degrees. First I logged 5 minutes of GC control, then I switched to PID control for 5 minutes. The red line is the switch.
www.mikepetro.org/coffee/images/GC/control.jpg
Typical process control response, the deviation error of each cycle is half the amplitude of the previous cycle until it reaches setpoint, it should stabilize within 3 cycles if the PID tuning is right.

Next step is to run it under full load. Might have to tune the PID a bit, but I except similar results. Wish I had some dirt cheap trash beans for this phase of testing, then I could do some sacrificial roasts and really push the limits of what this machine can do and what can be done with profiles. The machines maximum Rate Of Rise (ROR) is a value that is of particular interest to me.

More to come.....
Mike Petro
Martinsville, VA
--------------------------------
Lady Silvia, Stepless Rocky, Gene Cafe
http://mikepetro.org/coffee/
 
Jump to Forum:

Similar Threads

Thread Forum Replies Last Post
Gene Cafe Lifespan Gene Cafe Roaster 18 12/17/2021 4:10 PM
Opinions requested on the Gene Cafe CBR-101 Gene Cafe Roaster 3 10/21/2021 9:56 AM
Gene Cafe Artisan profile Gene Cafe Roaster 2 10/09/2021 6:58 AM
Looking to upgrade Gene Cafe WHAT SHOULD I BUY OR MAKE 7 09/24/2020 3:05 PM
Cleaning the roast chamber on a Gene Cafe Gene Cafe Roaster 4 06/29/2020 7:36 AM
Homeroasters Association Logo, and all Content, Images, and Icons © 2005-2016 Homeroasters Association - Logos are the property of their respective owners.
Powered by PHP-Fusion Copyright © 2022 PHP-Fusion Inc
Released as free software without warranties under GNU Affero GPL v3
Designed with by NetriX