Heat Staking Machine and Paint Flow Controllers for VividInc

Just a couple of short notes:
Firstly, I'm happy to report that Thomas Nguyen of VividInc was able to get the Heat Staking Machine working, the project that I reported on back in October 2019. He did have to make a small change in my design to select a higher current solid state relay but otherwise my design suggestions worked out just fine. I didn't visit Thomas to work on the system as he didn't need my help to complete the project. Now he will be able to use the system to make the clamshell covers for electric vehicle batteries.

Second, Thomas told me that they will need more Paint Flow Controllers at VividInc. Vivid's workflow had been cut back after their power problem and small fire at their Santa Clara facility but now are recovering and needing controllers for their work. I had reported that I had assembled three new systems and also that I had improved the response time of the systems. Soon I will be able to deliver them.

So, good news for Thomas that he now has the Heat Staking Machine working, good news for Vivid that they are recovering from their power problem at their facility and good news for me that I will get my assembled controllers out of storage in our garage and deliver them to Thomas after I check them out.

Here are a couple of photos of the latest Paint Flow Controller prototypes:

Controller Front View.

Controller Wired.

Controller in Test Mode.

I've written and published a book, "I am an Electrical Engineer"

My book is available at Amazon.

I wrote the book at the instigation of my son John William Baker. I frequently would meet my son for lunch and would regale him with stories of my projects at the U.S. Naval Research Laboratory (NRL) and the Lawrence LIvermore National Lab (LRL, LLNL) and my consulting projects and he suggested that I write this book. My son is a software engineer specializing in Java, currently working nearby in Pleasanton, Calif. 

My handwaving descriptions were inadequate to really describe my projects and he insisted that I put them to words. I had just finished up a project for LLNL and had time to put words and pictures to paper. 

I hope that my book will be interesting for technical minded readers with the many projects of greatly varying subject and technologies ranging from electromagnetic wave technology to digital technology to software and to Fuzzy Logic for control systems. The book is available in both Kindle format and paperback at Amazon.

Among my projects was development of an analog computer to control a shipboard antenna at NRL for our Moon Relay Communication project. 

Moon Relay antenna on U.S.S. Georgetown

Other NRL projects included our antenna experimental study at the FAA station on Molokai, HI. 

Then I wrote about my LLNL work at the Nevada Test Site during the atomic testing program, 

A typical atomic test installation.

the Liquefied Gaseous Fuels system development and testing at China Lake, 

Installing a Weather System at the Liquefied Gaseous Fuels Project.

my work in the Laser Department during the Atomic Vapor Laser Isotope Separation (AVLIS) Project, LLNL's Uranium separation and refinement project.

Laser tech aligning a Copper Vapor Laser in AVLIS

and then my work in the Global Security Department including work for the U.S. Border Patrol, our HIPROTECT system, and work at Fort Benning, an archaeological protection system for the abandoned U.S. Indian village at Yuchi Town. 

I included my consulting projects including a coin mechanism development project and several Fuzzy Logic paint flow control system projects. 

Paint Flow Controller operating in test mode.

I was lucky to have received a terrific EE education at the University of California, Berkeley under such luminaries as J. R. Whinnery and Victor Rumsey and then to have gone on to work with many terrific coworkers and bosses. Hopefully I have given adequate credit to them all. So Happy Reading!

Heat Staking Machine

In our office with Heide's poppy painting.

I contacted my buddy Thomas Nguyen in March to find out what was going on with him and things and at Vivid-Inc. Vivid was looking for a new location for their facility since they had had an electrical problem and small fire at their Santa Clara location.

Thomas said he was starting on a new project, a Heat Staking Machine for emplacing inserts such as those from Penn Engineering into plastic plates that would cover batteries, like clam-shells, for an electric car company. I don't know which electric car company but I can guess! Thomas gave me some diagrams of the clam-shells and said he wanted to heat metal plates that would hold the inserts up to 250 to 300 degrees F.

I first located a small pneumatic machine listed on the ebay site that might be a good place to start

Pneumatic machine listed on ebay.com.

and then worked on the project for several weeks, just a day a week, coming up with suggested heat strip designs using heaters such as those from Grainger. I selected 240 VAC heat strips to minimize the AC current to the heaters but then later on realized that they had only 120 VAC power. Once Thomas was satisfied with the heat strip designs, I began searching for controllers and other components.

Tempco Strip Heater.

I selected a temperature controller, the Inkbird PID controller plus a solid state relay, a heat sink, and a thermocouple. Thomas would need to purchase a controller for each heat staking machine.

Inkbird Temperature Controller, solid state relay, heat sink and thermocouple.

I suggested that I design and have Protocase build an enclosure or L-Bracket to house the temperature controllers etc as I did for the paint flow control systems. And I selected and recommended a Square D distribution box with appropriate circuit breakers to accommodate the power required for the heaters.

Thomas designed and built a large roll-around pneumatic press for the heat staking machine. He ordered the Strip Heaters, the Inkbird controllers and other components and then began working on assembling the whole system. I need to check back with him and see how he's doing.

Fuzzy Logic and Paint Flow Revisited

That's me in our office with my wife Heide's painting in the background.

Back in 2016 I was very hurried in my paint flow control development, trying to get the controllers ready for my cohort Thomas Nguyen to install in a painting facility in China that painted the covers of the Comcast remote control units. I ran out of time to optimize the Python control system code. However, I did a little "trick" in the software to speed up the response. The result was that the controllers worked well to control the pumps and paint guns. My trick was as follows: first the controllers started up in Manual Mode where the overall paint flow system gain could be determined, that is, the controller output value to the paint pump required to provide an adequate paint flow. Then that gain value was used to initialize the controller output in Automatic Mode and the controller would be switched to Automatic Mode. In Automatic mode, the controller output would be controlled further in closed-loop mode. The trick allowed the controller to quickly get the paint flow to the required Setpoint value and the controller would stabilize the paint flow. This trick got around the slow response of the Fuzzy Logic Algorithm (FLA) that I had developed. See the chart below. Recently I reconsidered and began to believe that I could improve the FLA so that my speed-up trick wouldn't be needed.

Original Paint Flow Control simulated very slow response.

The chart above shows that the paint flow reaches the setpoint in about 25 seconds. So I decided to see if I could improve the FLA to speed up the response. I had used Octave (Octave) along with L. Markowsky's Fuzzy Logic Toolkit to generate a Look-Up-Table or LUT that I then used in my Python code to produce the output signal for the paint flow pump. 

The LUT that I had previously generated is shown below.

Coarse LUT, August 23, 2015

The Error axis is in the forefront of the chart. The Error_Dot or derivative of the Error is shown along the right-hand side of the chart. The 3-D graph is not perfectly smooth and so I decided to try revising the Fuzzy Rules and compute a new LUT.

The August 2015 LUT worked pretty well but I decided to rework the Fuzzy Rules from the original 25 to 49 rules, then as before I used Octave and Markowsky's Fuzzy Logic Toolkit to produce a new LUT, shown below.

LUT, January 31, 2018

However, although the revised LUT shown above is a little smoother than the August, 2015 LUT, it did not to improve the simulated controller response.

Consequently, I decided to tweak my Python code and discovered that in my haste I had mistakenly set the output_scale_factor or gain constant to 6.0 when, for large errors > 200 the output_scale_factor should have been 250.0! So drastically increasing the output_scale_factor from 6.0 to 250.0 greatly improved the simulated system response. Below is the simulated response for the corrected output_scale_factor.

New Python code response, system time constant 2.0 seconds.

So problem solved, easily, and the controller response should be greatly improved. Next to load the software into my Beaglebone Black microcomputer and give it a try.

The Paint Flow Control Project and the Satellite Comm Project, July 2, 2017

Here I am with my wife Heide's painting in the background

Paint Flow Control Project: 
No order has been forthcoming from the Chinese painting facility for additional controllers but I did get the go-ahead to build three controllers from Thomas Nguyen of VividInc (VividInc). Thomas will use them at Vivid for in-house systems, painting projects for their customers. Vivid provides various industrial coating for many customers in their Santa Clara, CA facility.

Vivid will purchase the expensive parts including the paint flow meter equipment, the pumps and the I/P transducers for controlling the pumps. The flow meter consists of the flow sensor, a spur gear positive displacement sensor and the AW Lake MX9000 flow display unit (MX9000) which we have employed in the systems delivered to the Chinese facility. The I/P transducer is an Omega IP 610 (Omega IP610) that we can control with the Beaglebone Black microcomputer and my Analog board as in previous control systems. I am purchasing the Beaglebone Black micros (BBB), the LCD's (4D Systems LCD), the bare Real-Time-Clock and Analog boards from OSH Park (OSH Park) and the components for the the boards. To get a highly stable clock module, I am using the DS3231 from AdaFruit (DS3231) on my Real-Time-Clock board. OSH Park has provided other boards for me and their boards have always been first-class and reasonably priced. In addition, OSH Park software is compatible with my Kicad (Kicad) designs, making the ordering simple. My Analog board uses the PWM output from the Beaglebone Black board, low-pass filtering the PWM signal and converting to a 0-20mA signal to drive the I/P converter.

The Python code with my Fuzzy Logic control algorithm has worked perfectly in the Chinese facility. No modifications have been required or seemed useful. So I am simply duplicating the code for these new systems.

I ordered the various parts, mostly from Digikey (Digikey), Mouser (Mouser) and MCM Electronics (MCM Electronics) and have begun assembling them into the enclosure. I redesigned the enclosure, actually an L-Bracket, and ordered three from Protocase (Protocase). Protocase promptly manufactured and delivered them perfectly according to my design. I made a slight mistake on the hole positions for the LCD's but was able to elongate the holes a little to make the LCD's fit satisfactorily. Here are photos of a partly assembled unit, lacking the AW MX9000 flow meter display and the Omega IP610 converter that haven't yet arrived.

Front view of partially assembled controller unit.

Rear view of partially assembled controller unit.

I plan to get Noemi Fortes to assemble the controller units including loading the Analog and Real-Time-Clock printed circuit boards. Noemi is an old colleague from my days at the Lawrence Livermore National Laboratory and always does perfect work.

So far I have assembled the Real-Time-Clock boards and loaded the software on the BBB boards, All seems well so far and hopefully the LCD's and MX9000's will arrive soon. Then I will get the parts to Noemi to assemble the three systems.

Update on Satellite Communication Project for the LLNL Smart Sampler Systems:
I finished developing the code for the Smart Samplers (Smart Sampler) in mid-March. Steve Hunter of LLNL found and repaired a couple of errors in my code, installed the code in one of the SS's and was able to get it working over the satellite system. I reported on this in my last post and so will not repeat the description. The LLNL Project Leader, Charles Carrigan, was happy that the system worked so well but wanted to add several signals, some controls and some monitoring. Steve provided me an updated list of signals and I added code to implement the additions, providing new code to Steve by the end of March. Also Mike Saltsman (Mike Saltsman) of Purestream Technology provided a new web browser code. At first the updated systems seemed to work just fine but some glitches have shown up since then and Steve is trying to work them out. Steve thinks the problems are timing related with the satellite modem. Hopefully he will be able to work out the problems with Mike's help.

That's all for now.
John

Still waiting for go-ahead on Paint Flow Controllers but other work on a Satellite Communication Project

That's me again with painting by Heide in background.

Still no go-ahead on the Paint Flow Control Project. It had looked like UEI (Universal Electronics) was planning to get more controllers installed in the Chinese painting facility. The facility is in Guangzhou, fairly close to Hong Kong in the south of China. That's where UEI gets many of their remote control widgets painted. They were having QC problems, so UEI instigated the development project. I was working with Thomas Nguyen of VividInc (Vivid Inc.) on the controllers. We developed a control system using a Beaglebone Black micro with a 4D Systems touchscreen LCD and an analog interface board that I designed and developed. I programmed the BBB in Python 2.7. Thomas and I tested the system at the VividInc facility in Santa Clara, then Thomas installed four controllers in the Guangzhou facility. I have described all this in previous posts. During the installation of the controllers, Thomas also made several corrections/improvements to the painting systems such as straightening out the kinks in the paint flow tubing and installing filters to clean up the paint. The overall result was a halving of the reject rate and we had thought that the Chinese facility would go ahead to install many more controllers, maybe up to 24 or so. However, it turns out that they are unconvinced that the controllers are needed, instead suspecting that Thomas' fixes did the trick to improve the reject rate. So we are still on hold, awaiting a decision from the Chinese painting company.

Meanwhile, and actually a few months ago, Steve Hunter of LLNL (LLNL) contacted me about working on software for the Smart Sampler (Smart Sampler), software to enable satellite communication. I was happy to be working on it again and accepted an offer from Akima (Akima), the contracting company for LLNL. While employed at LLNL, I had originally programmed the systems in National Instruments LabWindows (LW/CVI). They were originally called Gas Samplers. Steve has since modified the code to add some features but he was too busy to add the software to effect satellite communications and so called upon me to help out.

I have been working almost entirely at home, using my old Dell Inspiron 3800 to simulate the Smart Sampler systems. The hardware is considerably different from the real Smart Samplers, without all the serial interfaces and sensors etc, but it uses the Windows XP operating system, similar to that employed in the Smart Samplers, so was a good surrogate. Steve and Charles Carrigan had set up a deal with ORBCOMM (ORBCOMM) for the satellite communication system and I worked with VitalVu on the software, mainly with Michael Saltsman of Vitalvu (VitalVu). The ORBCOMM Satellite Terminal (transceiver) looks like a Modbus Master and Michael suggested use of modbus

SkyWave IDP680 Satellite Transceiver

poll from modbustools (Modbus Poll) to get the communication working. I started with Modbus Slave code from Github (Arduino Modbus Code), adding much to the code and considerably rewriting it since the code was written for an Arduino micro. Modbus Poll worked out very well for testing my code and verifying that it was working properly.

Anyhow, with much help from Michael Saltsman of Vitalvu, Michael wrote special code for the Satellite Terminals to match up with our Smart Sampler data and also provided us with a website

VitalVu Liquistics webpage for Smart Sampler Project

that displays the data and enables us to turn on/off various systems and change the operating mode of the Smart Sampler. I got my LabWindows code fully operational on my XP PC. I set a Satellite Transceiver out on our driveway at home and got communication working to the ORBCOMM satellite and the Liquistics webpage.

Satellite Terminal on Driveway

Now it's time for Steve Hunter to recompile the code for the Smart Sampler hardware configuration and install it on a real Smart Sampler. Then real testing can commence. I expect Steve will get that done soon, and hopefully he'll have other work for me to do for LLNL.

Paint Flow Control Project, Chapter 3.27, Awaiting Go-Ahead

I've been super busy on a satellite communication project for the past couple of months. It's not all operational yet and I'll report on the project later. Meanwhile, the Paint Flow Control Project has been in limbo, awaiting a go-ahead from Vivid Inc (VividInc). Thomas Nguyen of Vivid tells me that they have a P.O. from the Chinese company to build a couple more controllers. The Chinese company's CEO is not convinced that it's worthwhile to outfit their facility with the controllers but apparently UEI (UEI) has insisted on installing another two controllers to complete a single painting system that requires four controllers for four paint guns. The full complement of Paint Flow controllers should be sufficient to prove or disprove the value of the controllers. So far, the installation of two controllers has improved the painting reject rate of 15% to 12%. It doesn't sound like a big improvement but since they will be painting hundreds of thousands of remote control units, the reject rate improvement will be a big money saving and should easily pay for the paint flow controllers. Hopefully Thomas will give me the go-ahead soon.

I haven't done much work on the Paint Flow Control project, just designing a simple board for the Real Time Clock. I've been using the Adafruit DS3231 (Adafruit DS3231) module, mounting it on a proto board from Osh Park (Osh Park), their "AKA mod of Adafruit BBB Proto Cape copy." The Adafruit DS3231 has worked well in our controller systems, being easy to implement for the Beaglebone Black and quite accurate. Although it is pretty easy to mount the DS3231 module on the Osh Park proto board and wire it up, I wanted to reduce the small effort required to wire up the DS3231 and decided to design a simple board for the DS3231. The board layout is shown below. I used Kicad (Kicad) as before, finding that it does the job for me and now Osh Park can take my Kicad pcb file directly and manufacture the boards.

Beaglebone Black Real-Time-Clock Cape Board.

Osh Park's quote for three of the boards is $30.15 and so I expect to get a few of the boards for the Paint Flow Controllers as soon as I get the go-ahead from Thomas and also I'll order a few of the redesigned Analog I/O boards from Osh Park. That's all for now.