Jump to content

RA5040

Members
  • Content Count

    89
  • Joined

  • Last visited

Community Reputation

31 Excellent

About RA5040

  • Rank
    Tamper Master

Your Profile

  • Location
    Slane, Ireland in the Boyne Valley near Newgrange (and the Slane concert)
  • Interests
    Painting, photography, coffee, hiking ...
  • Occupation
    Currently painter and photographer
  1. If you would be interested in checking out the current version of Gene controller/logger software, you can get it here: http://www.irelandupclose.com/customer/coffeeforums/Gene.zip Unpack the file. Then to run it open a command window in Windows, or terminal window in OSX/Linux. Run the program by typing in: java -jar Gene.jar You don't need any hardware ... the software currently simulates the hardware. You must first select a port at the top and click on Connect. Any com port will do. You can then click on Set Temp (to send temperature value to simulated Gene ... value will be shown in terminal window). Then click on Start to run a simulated roast (the data is in data.txt) Either during or before you click on Start you can Load Profile. Navigate to Profile.txt. This will show the profile (from a previous roast normally). Then if you want to you can click on Set Profile and this will start to send temperature settings to the Gene (terminal window for now). If you want to override the profile, click on Set Temp. and this will send the temperature setting to the Gene. You can restart the profile by clicking on Set Profile again. To end the roast click on Emergency Stop. Here's a quick video showing the program running: http://www.irelandupclose.com/customer/coffeeforums/Gene.mp4 Have a look at the folder: you will see that the chart and profile for this roast have been saved with a date/time names.
  2. If you would be interested in checking out the current version of Gene controller/logger software, you can get it here: http://www.irelandupclose.com/customer/coffeeforums/Gene.zip Unpack the file. Then to run it open a command window in Windows, or terminal window in OSX/Linux. Run the program by typing in: java -jar Gene.jar You don't need any hardware ... the software currently simulates the hardware. You must first select a port at the top and click on Connect. Any com port will do. You can then click on Set Temp (to send temperature value to simulated Gene ... value will be shown in terminal window). Then click on Start to run a simulated roast (the data is in data.txt) Either during or before you click on Start you can Load Profile. Navigate to Profile.txt. This will show the profile (from a previous roast normally). Then if you want to you can click on Set Profile and this will start to send temperature settings to the Gene (terminal window for now). If you want to override the profile, click on Set Temp. and this will send the temperature setting to the Gene. You can restart the profile by clicking on Set Profile again. To end the roast click on Emergency Stop. Here's a quick video showing the program running: http://www.irelandupclose.com/customer/coffeeforums/Gene.mp4 Have a look at the folder: you will see that the chart and profile for this roast have been saved with a date/time names.
  3. My initial results are as follows: - vacuum packing definitely keeps the beans fresh (don't know for how long at this point) - the vacuum-packed beans stored at room temperature degass a bit (varies depending on beans and roast level: Yirgachefe roasted light appears not to degass at all; Colombian Supremo roasted medium degasses a bit; Colombian Supremo roasted dark degasses quite a bit (about double the volume of the beans) - the vacuum-packed beans stored in the freezer do not degass, or only very little - I've only compared one set of Colombian Supremo stored at room temperature to the same beans stored in the freezer: the frozen beans had a more fragrant smell and taste initially, but after a day the frozen and room-temperature beans seemed pretty identical - Darker roast beans stored at room temperature show oil; the frozen beans do not (until they have been defrosted, then same as room temperature beans). - The frozen beans appear to have some structural change to them (I've seen this on two sets of beans). This is particularly evident for espresso: to get the same grind I've found that with my grinder I need to set the fine setting 2 to 3 clicks up (i.e coarser) for the frozen beans. So at this point I would reckon that vacuum packed is good. Whether to store at room temperature on in the freezer is unclear (although I don't like the 'structural change').
  4. This is more like the final version of the Gene Controller/Logger: http://www.irelandupclose.com/customer/coffeeforums/roasterLogExample2.jpg I've added 2nd Crack start and end, and also the instant temperature readings (at the bottom of the chart). I've also moved the control panel to the right (as I'm right-handed). Robert
  5. I'm making some progress. Here is a first crack at the controller app (written in Java ... so it will run on Windows, Linux or Mac, but not on tablets or phones): http://www.irelandupclose.com/customer/coffeeforums/roasterLogExample1.jpg The data is just made up as I don't yet have the Gene hardware available (parts on order though). What I've implemented so far is as follows: - Connection drop down list and button at the top - Set target temperature (same as Gene) - Set end time (same as Gene) - Load Profile: this allows the user to load a previous run and either have it just for reference, or instruct the roaster to try to match it (Set Profile) - Override the set profile by clicking on Set Temp. - Restart the set profile by clicking on Set Profile - Start/Cool/Emergency Stop - Emergency Stop can be restarted (to allow for dumping of beans and either doing a new roast or running the Cool cycle) - Click on 1C Start and 1C End to set markers for start and end of 1st crack (I'll add 2C Start) - Save current run as a profile that can be run, and also save chart as a jpg. - Errors from the Gene will be reported (and appropriate action taken). As you can see, I'm intending to monitor input air temperature, exhaust air temperature (from current Gene probes) and bean mass temperature (probe in the drum). As it's possible to determine the drum position from the Gene hall-effect sensor, measuring BMT should be fairly good. I will most likely use an ESP32 for the main controller as it's much faster than the Atmega328. I intend to phase control the heater. ... still, all some months away yet! Cheers Robert
  6. Hi, yes, several people have done this ... nothing new Here's another one: http://coffeesnobs.com.au/home-roasting-tips-tricks-ideas/48343-gene-automation-mod-lcd-version.html I just wanted to find out how to do this and I think I pretty well know how to now. But I'm not in any rush to do it: I've only had the Gene for a couple of months and I want to get to know it well as it is before carving it up. I'm quite impressed with the Gene's roast quality, I have to say, and I doubt that this sort of mod would make things fundamentally better. Perhaps it would make it easier to repeat roasts more consistently ... but then again, part of the charm of roasting at home is the (somewhat unpredictable) variations in the roasts. So at the end of the day I think I will do the mod, but more for fun than anything else. But I would be more than happy to share everything that I do, if there is any interest. I will do up proper printed cuircuit boards and I can always get a few extra ones done for anyone who would like one. Cheers Robert
  7. Not having implemented any of this yet, my input is of debatable value. Having said that, I have thought about it all quite a bit, and my own feeling is that the dimmer mod on its own has limited value (I wouldn't be bothered doing it). I think the BMT probe is probably worth doing as it would give a useful graphical view of the roast and would help to run roasts based on stored results (at least to some extent). But I do think that the best approach is to replace the Gene electronics (with or without the BMT probe). This would give complete control over the Gene, it would allow both the input air temperature and exhaust air temperature to be plotted and stored, so that roast profiles could be stored and new roasts run on these profiles; and the Gene heater could be controlled much more accurately with PID phase control. The drum position would be known as the Gene already has a hall-effect sensor, so if a BMT probe was added (which would certainly be a good addition) it would be easier to get a good BMT reading (since knowing the drum position would help in taking the BMT reading). Then the environment temperature could be worked out, perhaps as an average of the input and exhaust air temperatures, and this would allow for a more accurate computation of the bean mass temperature. I've had a look at the Gene electronics and there really isn't much to it: motor, fan and heater control and that's about it. Robert
  8. OK, well it should be quite easy to connect to the Gene over MQTT in Java, and it is easy to do up a Java GUI with graphs using JFreeChart. Here's the result of a bit of messing with Java/JFreeChart, with simulated values in this case: What I'm planning to do is to develop this app so that it has good functionality (for example, being able to store a roast profile and use it, as is or modified, to run a roast). I will then install Ubuntu on an old Android tablet I have and run the app on that (I don't particularly want to use my Apple laptop, but that would also be an option). I will use USB or bluetooth, but I don't see any reason why MQTT would not be fine (Java has MQTT libraries). Cheers Robert
  9. And this seems to fit the bill for MQTT (but I've only had a quick scan through ): https://www.hivemq.com/blog/how-to-get-started-with-mqtt It wouldn't seem too hard to fit this and the JFreeChart together for a pretty nice system. But I can't quite see why you would want to use MQTT to control a Gene, to be honest ... I think USB or bluetooth would be preferable (and simpler!). Robert
  10. Here is a really excellent tutorial on using JFreeChart with a connection to an Arduino. As you can see, it would be very easy to add additional temperature sensors, set point for temperature PID (or implement the PID in Java) etc. It's certainly an alternative to roastlogger, and one that we would have complete control over. Robert
  11. JFreeChart is normally used from within a Java app so there's no reason why MQTT data can't be used alongside it. I'll have a look as I also need something like this, and unfortunately roastlogger seems a bit closed (although it may be possible to adapt it? Rob1 would know more about this than me).
  12. How about using JFreeChart? Open source for Mac, Windows, Linux.
  13. Here's a sketch of what I intend to do ... that is, replace the Gene electronics completely. It seems like the logical thing to do, even though it's a bit more work (but not much more than doing the power mod & adding a BMT sensor). The advantages of replacing the electronics are that it gives additional control/information: for instance the inlet and exhaust thermocouples and the drum position sensor become available; also bypassing the power to the heater is no longer a safety issue as the new electronics can shut down the system if temperatures go too high (anywhere). I will probably end up using one of the Arduino/Adafruit boards but as it will be necessary to build a circuit board it may end up not being more difficult to use the microcontroller directly. Robert
  14. What you could do is to put a hall-effect sensor on your circuit, with a magnet on a fixed surface of the Gene - that would give you a fixed position of the drum for every rotation.
  15. Yes, I can see that the rate of cooling is reducing ... however as long as the temperature is dropping, the rate of rise should be negative. It should be computed as (current reading - last reading)/(time period), so as long as the current reading is less than the last reading the ROR should be -ve. Yes, the bean mass temperature not being consistent could be a reason. Another reason could be that with a single probe it will be very difficult to get a reading of both the air and bean temperatures. A slight variation in reading is all that's needed to make the differential very jaggedy. It would probably be better to use a weighted average of the last 3 readings, say (or just use the smoothed curve) Thanks for that ... I'll see if I can find the link. Cheers Robert
×
×
  • Create New...