Paint Control Project, the Saga Continues

This is me in my Lab in front of a painting by my wife Heide.

In my last post, I reported that Universal Electronics Inc, UEI (http://www.uei.com/), and Thomas Nguyen of VividInc (http://www.vividinc.com/) in Santa Clara are interested in manufacturing my controller board with the hope of selling Paint Flow Control Systems (PFCS) in Asia. UEI is in the consumer electronics business and makes electronics gadgets in Asia. They recognize that Asia may be a good place to sell the PFCS because so much electronics manufacturing is done in Asia. UEI and Vivid are working on setting up an R&D facility for the Systems at VividInc. A primary use of the PFCS is for control of copper-based paint used for RFI shielding of electronics units. One of the main components of the PFCS is the flow meter to measure the paint flow and UEI began to search for a low-cost flow meter for the R&D facility and for possible sales with the system. I also started looking for a flow meter. In the past, very expensive flow meters called Coriolis flow meters were used (https://en.wikipedia.org/wiki/Mass_flow_meter). The Coriolis flow meters are mass flow instruments and have several advantages including accurate measurement of paint at low flow rates, having no internal parts to get clogged up or damaged by the abrasiveness of the copper-based paint that is used for RFI purposes. however, like I say, the Coriolis flow meters are quite expensive at approximately $10K each and the R&D facility will need four of the flow meters. So I hoped to find a flow meter that would handle the viscosity of the paint and survive the abrasiveness of the copper powder in the paint. Other types or flow meters include the magnetic flow meters and the positive displacement meters that generally use oval gear system to measure flow. The magnetic flow meters are fairly accurate, < 1%, and are appropriate for the copper-based paint since they don't have any moving parts. However, the flow rate of our system is very low at approximately 20 to 40 mL/min, too low for the magnetic flow meters. However, I found at least two positive displacement flow meters that may be able to survive the abrasiveness of the paint: both manufactured in Australia. The oval gear positive displacement flow meters are manufactured by Trimec and MacNaught. Of course there may be other similar flow meters but these are the ones that I have found so far in my searches. They are much lower cost than the Coriolis flow meters and are sufficiently accurate for our process, < 1%. However, Trimec and MacNaught cannot guarantee that their flow meters will survive the copper and water-based paint, but it seems worthwhile to give them a try and I am checking the prices of these flow meters in hopes that they will do the job for us. I've been working with Colin Willard of Instrumart (https://www.instrumart.com/). Colin suggested the MacNaught flow meter and has suggested one from AW Gear Meters (http://www.awgearmeters.com/main.html), the SLG spur gear flow meter that is intended for use in the automobile industry for paint flow measurement. It is more expensive than the MacNaught flow meter but may be more appropriate for our use, so I have asked Colin to check into the AW SLG meter to see if the SLG will handle the water-based paint.

Another aspect of the project is 1) purchasing microcomputer boards from Digi Inc, their BL2600, and programming them with my Fuzzy Logic code or 2) manufacturing my controller board with its Microchip dsPIC33FJ256GP710A microcontroller chip. I have some experience with the BL2600 board for the SilkScreen Project that I reported on earlier and it appears to be perfectly adequate for the PFCS. However, there seems to be no way to protect the code from piracy. The Microchip dsPIC chip can be protected from piracy and so I prefer to manufacture my controller board. However, that is still to be determined. UEI may prefer to use the commercially available BL2600 board and hope for the best regarding piracy. Anyhow, I decided to get some prices on manufacturing my controller board and found that SFCircuits provided a preliminary estimate that they can manufacture ten boards for $400 and also assemble the ten boards (solder the IC's and other components on the boards for $140 per board. For the first version of the board, I had my old friend from Lawrence Livermore Lab, Noemi Fortes, assemble the boards. Noemi did an absolutely perfect job and so I would like to get her to do the assembly work on the next set of boards. However, if we go ahead with the board manufacture, UEI may prefer to get a PCB assembly house to do the work. So we will see how this plays out. Also, of course, the whole project is still up in the air as UEI and Vivid have to make a deal on setting up the R&D facility at Vivid and I still want them to make a deal with me on the controller. So I'm hopeful that the details can be worked out and we can keep the project rolling.