I recently got a call querying the details of how we automated the extraction process on the WolfWurx Mk VA2’s, but the story actually predates WolfWurx and starts with the Mk II Terpenator and the Mk VA.
We automated the extraction sequence on the Mk II using a programable controller and adjustable pressure switches. We subsequently replaced that PC with a custom control system that we developed engineering the Mk VA.
The control boards for both the Mk VA and Mk VA2’s were custom built by Pacific Semiconductor Inc here in River City, based on a chip, which we built our control panel around.
The principle Bob at Pacific Semiconductor Inc is semi-retired and can be reached at 503-274-2471. Semi means he is not working to stay alive, but for the enjoyment to keep his juices flowing, so is highly selective.
We were fortunate to be able to have our chip programmed by the same software engineer who programmed the Microsoft Pentium III & IV chips for Microsoft and it worked flawlessly at a considerable savings for that same capacity using an industrial PC.
I supplied the control logic to Bob at PSI and between him supplying the circuit boards and Brian providing the programming, I received the four programmed circuit boards ready for my son and I to mount and build our two control panels around.
Besides those furnished by PSI, I purchased other unique automation components at:
Automation Direct 1-800-633-0405 https://www.automationdirect.com/adc/Home/Home
Mar Vac Electronics 1 (800) 655-6686 https://marvac.com/
As you can see from our first Mk II control board below, the PC controls drive solid state Crydom style SSR’s and the three adjustable pressure switches in the first Mk II prototype were replaced with a single pressure transducer.
Mk II Terpenator control cabinet using PC and pressure switches
For the Mk VA and VA2 we replaced the commercial Programmable controller with our own custom programmed chip, which Pacific Semiconductor, Inc built the control panels around.
WolfWurx Mk VA2 automation controls using transducer
Ok, we’ve have logic and electronics that send a signal to an SSR, what does it activate? In this case I used pneumatics to operate the valves, so the SSR’s controlled three or four-way solenoid air valves, supplying pneumatic valve operators.
I used three way valves and operators with spring return on valves that I wanted to fail closed should we lose power. In this case, the liquid LPG injection valve and the valve isolating the vacuum pump from the system.
I used four way valves on the rest, where I was powering operators to both open and close.
I mounted the eleven (11) pneumatic solenoid valves on a common ½” copper manifold, with the pressure regulated 80 psi.
Note that the air manifold had a pressure relief valve to break at 85 psi, so as to limit how much pressure the Haskel pump could generate.
Mk VA2 Air logic manifold
Probably a good time to talk about NEMA 7 Class 1 Division 1 electrical requirements. Electrical requirements have changed since the Mk VA2’s, and there are a couple ways to meet the ones requiring C1D1 or C1D2 electrical.
One way is to use an explosion proof cabinet, or render it so with a compressed air kit, and explosion proof solenoids on the air valves.
Another is to locate them outside the room and plumb through to the C1D1 enclosure with a tube sheet like the Mk II installation below:
Mk II Air logic manifold and tube sheet pass through for the plumbing
The control panel and all electrical beyond micro amp thermocouple leads can be relocated to an adjacent room such as this manual Mk IVC installation.
Mk IVC Equipment room adjacent to C1C1 extraction booth.
I used 3 and 4-way pneumatic actuators on stainless ball valves with PTFE seats and packing that I obtained from Paramount Plumbing Supplies here in River City. That is also where I obtained the solenoid air valves.
The pneumatic operators not only provide a lot of operating force in a small package, but are C1D1.
WolfWurx Mk VA2 automated valves
After getting the hardware in place, we developed the following solutions to automation logic:
How long do you flood a column for?
A set time doesn’t work because of variables like LPG pressure, temperature, fuel mix and column packing density.
Our solution was to ask instead, how many column volumes of LPG we wish to pass through it and from which direction(s)?
We found through running manual systems that on average three volumes were adequate for most of our product and purposes, especially if the first two were from the bottom and the last was a rinse from the top. We designed our system volumes around that premise.
We built our system as a prompting program that was energized when the system was powered up. It first asked how long you wanted to flood for, and you answered by toggling a spring loaded three position programming switch up to add to time in 15 second increments, or down to reduce it.
We knew from experience that it took from about 1 minute 15 seconds to 1 minute 45 seconds to flood a 4” X 36” column from the bottom until it overflowed out the top, so typically we would set the timer at 1 minute 45 seconds.
To tune the machine to the exact pressure, temperature, gas mix, and packing parameters unique to the column being extracted, as soon as the LPG overflowed the top of the column, as observed through a sight glass, the spring loaded programming switch was toggled downward again, which reset the previously estimated flood times to the exact time.
If not reset to actual, the program proceeds as initially programmed.
The second question it asked when operator was programing flood time, was how many floods from bottom, followed by how many floods from the top. Both questions answered by toggling the programming switch.
Once those two parameters were programmed, you toggled the spring loaded start switch, and the system:
Started the vacuum pump and energized 3-way air solenoid valve, which supplied air to the operator on a ball valve, which opened the valves between the vacuum pump and the recovery chamber.
It also opened all valves required to open the valves needed to fully vacuum purge the entire system of residual atmosphere.
When the system vacuum level reached -29.9” Hg as seen by the pressure transducer voltage, all of the valves, except for the column vent valve, are closed, and then the vacuum pump was turned off.
With the vent valve at the top of the column remaining opened, the 3-way/spring return LPG injection valve is opened, allowing the LPG to start flowing into the bottom of the column and filling it.
I used a 3-way spring return operators on both the vacuum pump isolation valve and the LPG flood valve, so that should we lose power or air, those would automatically fail closed.
The recovery pump starts and the isolation valve between the pump and recovery pot is opened.
The system continues to flood until it full fills the number of bottom floods programmed, which it determines by how much time has passed.
If it took 1 minute 15 seconds to overflow the column, and you have two volumes from the bottom programmed, it will continue to flood until another 1 minute 15 seconds, or a total of 2 minutes 30 seconds has passed.
It then closes the bottom flood valve and top vent valve, after which it opens a valve supplying the LPG to the top of the column, and open the isolation/dump valve between the column and collection tank.
The system floods from the top for the number of volumes programmed, using programmed or the actual flood time reset during startup. IE, for 1 minute 15 seconds using the example above.
When the top rinse times out, the LPG injection valve closes and the recovery pump continues to run.
When the system has recovered to -15” Hg vacuum level, the isolation/dump valve between the column and the pot close, as does the column vent valve.
A valve opens that gives the recovery pump direct access to the column, and 150F column heat is turned on.
If the program is allowed to run its course, the system will recover to -22” Hg, at which time the recovery pump isolation valve closes and the pump is turned off.
Note that here again, the system can be reset by the toggle of a spring loaded switch, to recover to different vacuum levels, and here is a point at which visual assessment of the concentrate pool is necessary for artisan products like Cotton Candy.
The operator views the pool through a lighted sight glass to determine when it reaches the correct “blurpy” stage to end recovery and inflate the concentrate using the high vacuum pump.
The high vacuum pump starts and the vacuum pump isolation valve is opened. The system recovers the column and pot separately to the level of -29.5” Hg.
When -29.5” Hg is reached, the vacuum isolation pump valve is closed and the N2 backfill valve opened until the system reaches atmospheric pressure (0 psi gauge), at which point the N2 back fill valve closes, followed by the vacuum pump isolation valve opening again, allowing the pump to pull the system back down to -29.5” Hg.
When the system reaches -29.5” Hg the second time, the vacuum pump isolation valve closes and the vacuum pump is turned off.
The system then opens the N2 valve to back fill again to atmospheric pressure (0 gauge), at which point the system shuts down and the system resets for the next run.