Phase 2 brewery build progress shots
- Thread starter kladue
- Start date
Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:
OP
OP
There are 3 proportional control valves for liquid flows and 2 mass flow controllers for LP flow control to burners. The liquid flow valves control water flow through flash boiler, wort flow through pump, and coolant flow to CFC wort chiller, and the mass flow controllers control gas flow to boiler and boil kettle burners. There are 2 sets of PID loop parameters for water flow, wort flow, gas flow to boiler, one loop for steam injection wort heating, and one loop for coolant flow. The loop tuning parameters are set and stored in database tables by using the loop tuning portion of the software package, and are selected and loaded into loops as needed during system operation. This allows one to tune for sparge water, strike water, wort heating, wort chilling, and wort recirculation to get best PID loop performance by swapping optimal parameters into loops for each condition as needed.
OP
OP
Getting closer to automatic run, have loaded and ran the java app on the panel pc and have begun the PID loop tuning needed for the automatic sequence operation. I should be able to wrap up the tuning tomorrow and test the malt handling system through dough in to see if mill, conveyor, and mixer work as designed. After dough in testing is done then it should be time for the first full automatic sequence run with 3 step mash sequence. Next up is insulating and jacketing the MLT, and boil kettle, then covering the cooling system tank and tubing so cooling system controls can be tuned for near freezing system coolant. Then it will be time to tear everything out and finish the inside of the brew shed and reinstall first control panel underneath 2nd control panel with touch screen to make room for automatic malt measuring system over grain conveyor hopper.
Sawdustguy
Well-Known Member
Wow Kevin, you are taking automation to a new dimension. Good show.
We expect pics of the first brew. 
Wow Kevin,
You never cease to amaze me! Can you elaborate on the automatic malt measuring system?
Sounds very intriguing. Have considered writng an article on the ultimate brewing system for BYO when you are finished?
Are you finished in Arizona and back home now?
You never cease to amaze me! Can you elaborate on the automatic malt measuring system?
Sounds very intriguing. Have considered writng an article on the ultimate brewing system for BYO when you are finished?
Are you finished in Arizona and back home now?
OP
OP
The approach for the malt measure system was to keep it simple by using slotted tube inside another tube to pass known amount each revolution. The control would be enable drive and count up to desired amount of rotations for an assumed volume. I have installed load cells and scale equipment from a couple different manufacturers and understand the complexity of the measuring systems so I chose to keep it simple and scalable. The largest problem is time to build it while home on R&R trips, control software will be an extension of the existing package running now that I work on nights and weekends to fill spare time. By next trip I hope to have the recipe formulation code and graphics completed so you can enter amounts and times and predict the results in the boil kettle and fermenter, control functions are currently tied to schedule information in current package.
Looks like I have commitments in Arizona until March 2010, might have additional work into 2011 if current bid is successful. So for this week it is do as much as possible while home then back to Arizona for another month.
Looks like I have commitments in Arizona until March 2010, might have additional work into 2011 if current bid is successful. So for this week it is do as much as possible while home then back to Arizona for another month.
OP
OP
System testing was a partial success, what works well on a 2.3 Gig processor has problems on a 500 Mhz 256 Meg panel pc. Did manage to get the calibration and some of the PID tuning done in the time available, still have to finish the boiler control loop tuning. All system components check out from the manual control screen, and the sequencer appears to work, just need to finish loop tuning now. Here are a few pictures of the testing on the panel Picasa Web Albums - Kevin - Panel Testing.
OP
OP
Here are some shots from the recipe generation and control program development for use on the touch screen panel pc that controls the hardware. This is linked to the system operating program and when you are done with recipe build and loading ingredients, you hit auto run button on screen and come back when it is done. Sorry for the pictures but Windows paint does not like the jpg format. http://picasaweb.google.com/kevin.ladue/RecipePanels#
Next part of the programming effort is the export/import of the recipe tables and graphing of the operating parameters during brewing. Maybe in a year I will get to actually spend more than a one day a month using the system.
Next part of the programming effort is the export/import of the recipe tables and graphing of the operating parameters during brewing. Maybe in a year I will get to actually spend more than a one day a month using the system.
Sawdustguy
Well-Known Member
Amazing stuff, can I come live at your house for few months?
That makes two of us.......:rockin:
OP
OP
After the software is completed in Java I am contemplating rewriting it in C++ for better memory management and reduced program code size( current compiled size in java is 5Mb). I currently have to do some interesting things to control memory use with the Java version when switching between frames with the quantity of bound variables on each pane so the code will work on the 256Mb panel Pc. When I run out of things to do for the new system I am contemplating a scaled down version of the software for use with the other hardware platforms. Not sure what folks would want or what would be needed for the standard 3 vessel brewing systems out there.
samc
Well-Known Member
I would think that if you are planning to sell your software for a total automation set up like you have you would need to create an entire package, hardware included. I doubt there are very many people who have a sophisticated system like yours for one touch brewing. I volunteer for a beta version, let me know when to pick it up?
OP
OP
I am not interested in the production of a package to compete with the existing system sellers, it was more of a something to do exercise to take up time. This is a hobby, I am not interested in making this a commercial enterprise as the liability that goes with the production and sale of equipment packages is not worth it. The first time some one was injured or property was damaged when using the product, it would most likely cost more than the entire venture was worth.
Updates?
How many beers have you brewed on this system?
I've shown some of the automation guys I work with systems like yours and they're blown away.
Shoot, I'm blown away. Very nice system. I only hope that you get to brew with it sometime, but as you've already alluded, this system seems more about the journey than the destination.
Cheers.
How many beers have you brewed on this system?
I've shown some of the automation guys I work with systems like yours and they're blown away.
Shoot, I'm blown away. Very nice system. I only hope that you get to brew with it sometime, but as you've already alluded, this system seems more about the journey than the destination.
Cheers.
OP
OP
Sadly I have not brewed a beer on this system yet,when I had finished construction in February the software was not ready. Since then I have been on the road in Arizona and have completed the control software but time on the rig is limited to about 8 hours once a month on the R&R trips home. The software has expanded to handle system hardware setup and calibration, monitoring for safety conditions, recipe calculation and control parameter calculation, PID loop tuning, and recipe save/retieve the database tables. Software currently runs on an Advantech touch screen panel pc which communicates with the opto 22 hardware over ethernet/RS-422 communication network. All that stands in the way of an automatic test run is the PID loop tuning on the panel pc, old tables were lost in software revisions that reconfigured the PID loop parameters for easier tuning and improved operation. I am still designing, purchasing hardware, and programming for additional features that I hope to build when I get done with this construction project next fall and return home for more than a weekend at a time.
GreenMonti
Well-Known Member
- Joined
- Nov 29, 2009
- Messages
- 1,268
- Reaction score
- 67
WOW. I am just now seeing this.
I don't think words can describe what you have built. Very nice! I would get lost standing in front of that thing.
I don't think words can describe what you have built. Very nice! I would get lost standing in front of that thing.
You are welcome to use my system as a basic hardware platform. I still seem to be having issues with my software
and your control screens are fantastic, even controlled by keyboard instead of touchscreen.
OP
OP
Beerthirty,sorry to hear that you are still having problems with the brewtroller board, i had hoped that a download of the current software had cured the problems. Looks like I need to purchase a Arduino board and start looking at a bit of programming to write the serial interface, then I should be able to make this application work as a front end. Just about done with this version of the software, just working out the recipe file save/ retrieve code issues with hiding and coloring button backgrounds.
trailblazer
Well-Known Member
if i was in your garage without your help about to brew a batch of beer, were would i even start? that is an amazing setup.
OP
OP
The system is relatively simple to operate, turn on the propane and oxygen tanks and go to recipe screen on touch screen.
Select batch size
Select BJCP style
Select fermentable type and enter quantity
Select mashing schedule
Select hops and enter quantity and boil time
Select adjuncts and enter quantity
Place fermentables in conveyor hopper
Place bagged hops on automatic drop positions
Connect carboy to CFC outlet
Hit Automatic sequence button on screen and the rest is automatic.
Here is a link to some of the control graphics, there have been improvments and additions since these shots have been takenhttp://picasaweb.google.com/kevin.ladue/NewGraphics#5411576901973950962
The strike water needed, strike water temperatures, sparge water needed, preboil and post boil gravities, color, IBU's in finished beer, and other parameters are calculated and displayed as you enter the ingredients. If you want to use a previous recipe you can select it from menu of recipe files and then place ingredients in appropriate hoppers, connect carboy and hit auto start button.
This system is an ongoing design, build and program exercise that helps to occupy my spare time while working away from home. One of these days I will get more than a weekend a month to work on the hardware and complete the PID loop tuning then start brewing again.
Select batch size
Select BJCP style
Select fermentable type and enter quantity
Select mashing schedule
Select hops and enter quantity and boil time
Select adjuncts and enter quantity
Place fermentables in conveyor hopper
Place bagged hops on automatic drop positions
Connect carboy to CFC outlet
Hit Automatic sequence button on screen and the rest is automatic.
Here is a link to some of the control graphics, there have been improvments and additions since these shots have been takenhttp://picasaweb.google.com/kevin.ladue/NewGraphics#5411576901973950962
The strike water needed, strike water temperatures, sparge water needed, preboil and post boil gravities, color, IBU's in finished beer, and other parameters are calculated and displayed as you enter the ingredients. If you want to use a previous recipe you can select it from menu of recipe files and then place ingredients in appropriate hoppers, connect carboy and hit auto start button.
This system is an ongoing design, build and program exercise that helps to occupy my spare time while working away from home. One of these days I will get more than a weekend a month to work on the hardware and complete the PID loop tuning then start brewing again.
Hi Kladue,
Thanks for the pointers yesterday, I'm now well and truly down the rabbit hole!
I think the programming language is fine. Maybe it's how you're using it (sorry, I mean no offense here). 5MB compiled does sound very large for Java and you mention that you're using a lot of bound variables which will also impact on the runtime efficiency.
Are you using an IDE to develop on (JBuilder etc) or a text editor and command line?
If you move the bound variables into objects, then instansiate them when needed, then the java virtual machine can deal with them much more efficiently and the garbage collector will dispose of them properly.
You can also improve efficiency by not creating all the panes at once, but instead do it when the user clicks on the tab.
OP
OP
The control program was built using Netbeans for the various classes and MySql for the numerous tables in use during operation. There are 2 frames in use, startup frame with 7 panes and the recipe frame with 6 panes. These are static and are instantiated once in an object where switching between the panes is handled as are all the other variables and points in use. There are 6 primary objects with over 38 associated classes, most are threaded for cpu load reduction, logic, or control looping, and 5 classes with SQL table generation and data storage routines. Most of the data manipulation and display is through customized Jtables and buttons for use on a touch screen platform. This makes for a lot of code not needed for non touch screen applications just to scroll and select items in the tables. The Jtables are filled from SQL tables and changes made to the data are written back into the SQL tables for later usage. With about 60 points in the system and typicaly 2-3 variables per point along with all the necesary control variables and table data variables the code has kind of evolved beyond what was originaly thought necessary. The last build reported 62 source files and 310 files total(most were icon graphics which are going away). This version is about half the size of the previous version and is shrinking as I work at removing Icon graphics for buttons and labels and resort to changing text and background colors when data changes. The other process underway is moving from individual labels for data presentation into custom Jtables to cut down on the total number of things on each pane.
OP
OP
About 30 to 35 hours each week to kill with learning Java and programming for last 10 months, with another 8-10 months to go until this job is done. With about 8 hours a month available to test on actual hardware, I have to finish the flow and heating tuning on the panel pc before autorun test is made. I built a simulation class so the sequencing and PID loops could be tested without hardware connection and any logic glitches could be fixed before committing any malt to the testing. With current hardware and software the brewing and fermenting temperature control are taken care of, enter recipe information and select fermentation temperature and deadband the rest is controlled by the system.
Next step is to build the hardware for the malt measuring system so one could enter recipe and have equipment measure out malts for batch. There are 2 approaches to measuring, known volume and cumulative weight, I have opted for the KISS method of known volume as it is easier to implement with digital control than a combination digital and analog control scheme.
Next step is to build the hardware for the malt measuring system so one could enter recipe and have equipment measure out malts for batch. There are 2 approaches to measuring, known volume and cumulative weight, I have opted for the KISS method of known volume as it is easier to implement with digital control than a combination digital and analog control scheme.
Sawdustguy
Well-Known Member
All I can say Kevin is "You da Man!"
OP
OP
I am still working on the software and have incorporated a data logging function with popup menu for choice of items to monitor. It saves the information in sql tables and csv files which are used by the auto scrolling graph popup that I built. The graphing packages I found were quite big compared to the roughly 40 lines of code needed to make the auto scroll graph work. Next up is to add auto ranging of Y scale and split Y scale for control and variable display for sub system operation. I hope to upload and tune next R&R trip in march so I can do a trial run with malt to verify system operation.
Boerderij_Kabouter
Well-Known Member
No dice on the video Kladue. It says it is a private video and I must accept a friend request before viewing.
OP
OP
Lets try this again
This is the current control program for the Phase 2 brewery, the gauges and buttons are relocated and reranged as needed to display analog values. Short automatic sequence run at the end, full auto run would be a couple hours. All that is needed for brew session is place malt bill in hopper, hop bags on drop hooks, and connect fermenter, all other functions are program controlled.
This is the current control program for the Phase 2 brewery, the gauges and buttons are relocated and reranged as needed to display analog values. Short automatic sequence run at the end, full auto run would be a couple hours. All that is needed for brew session is place malt bill in hopper, hop bags on drop hooks, and connect fermenter, all other functions are program controlled.
OP
OP
The wet sensors in water, mash, and boil are temperature and SG corrected based on measured vessel temperatures and estimated SG from recipe calculations. This program is built for operation on a 500MHZ 256 Meg memory touch screen panel pc, effort was made to keep memory and CPU loads minimized to allow reasonable operating response. Only initial pane is fixed, all others are dialog panes that popup when activated, and first pane is iconified to recover memory.
What hardware platform do you plan to use, change the I/O on this to match hardware and it should work for other systems.
What hardware platform do you plan to use, change the I/O on this to match hardware and it should work for other systems.
OP
OP
I paid about $10 -$25 each from ebay, pressure range is 0-2 Psi (0-54" water column). I have seen some for about $50 but if you are patient you should be able to find some cheaper.
OP
OP
The sensors are stainless steel wet sensors with a 2 wire 4-20MA loop powered interface. I will have to lookup the make and model numbers and let you know what they were. What is interesting is no one has tried to make a liquid isolator with saran wrap sandwiched between plastic flanges. Glue sensor on one side and connect wet pressure on other side, if there are no air leaks on sensor it should work well. We use same principal to isolate hazardous chemicals from pressure devices in industry, but I have not used them for this low a pressure range before.
With a negative offset value applied to signal, density correction formula for temperature compensation, and time scaled SG factor, the freescale sensors should be fairly close most of the time.
With a negative offset value applied to signal, density correction formula for temperature compensation, and time scaled SG factor, the freescale sensors should be fairly close most of the time.
I've had very good results with my sensor ports mounted in the bottom of the vessels and a tube running to the sensor mounted at the top of the vessel. I'm just curious if there are more robust sensors that I could mount directly in the bottom of the vessel and do away with the tubing alltogether.
Also how are you dealing with boilover? is your vessel just big enough to handle any amount of boilover foam?
Also how are you dealing with boilover? is your vessel just big enough to handle any amount of boilover foam?
OP
OP
I have chosen to mount the sensors on tubing away from the heat to help reduce the drift of the not so perfectly temperature compensated pressure transmitters. The water tank is a STI 0-2 PSI 2 wire pressure transmitter, the mash tun and boil kettle are Heisse 0-2 PSI 2 wire pressure transmitters, and false bottom presure drop differential pressure transmitter is a 2 wire 0-100" WC Mamac. I still cuise Ebay and elsewhere for parts for future system additions, and keep an eye out for suitable pressure and flow transmitters under $50 for folks interested in automation.
The trick passed to me from BeerThirty is the Simetecone drops which control foam, system also tracks temperature rise and reduces burner gas flow when temperature stops rising, and then it starts the hop addition timing for hop additions based on recipe entries.
The trick passed to me from BeerThirty is the Simetecone drops which control foam, system also tracks temperature rise and reduces burner gas flow when temperature stops rising, and then it starts the hop addition timing for hop additions based on recipe entries.
Similar threads
