While visiting Texas Instruments in February of 2019, marketing maven Gayle Bullock took me and analog guru Paul Grohe into the storage area where all of Bob Pease’s stuff was stored. The yearly Analog Aficionados party was that weekend, and Gayle thought we could give away some of the books and gizmos that were in storage.
Only a few minutes had passed when Grohe pulled a large copper-clad prototype fixture out of the pile and exclaimed, “Look at this, the LM317 voltage regulator prototype! (Fig. 1).” The LM317 adjustable voltage regular chip was a very popular IC designed by Bob Dobkin at National Semiconductor before he left to found Linear Technology, now part of Analog Devices. We assumed Pease’s hoarding instincts made him grab up this prototype, knowing its historical significance.
1. Texas Instruments application engineer Paul Grohe holds up an LM317 voltage regulator IC prototype likely designed by Bob Pease to improve the regulation accuracy over temperature.
I wrote Dobkin, and he replied, “In ancient days, I designed the LM317 and Pease designed the LM337 a couple of years later. He did not have anything to do with the 317 development. The 317 was designed in 1975 and introduced in 1976 if I remember correctly.”
The intriguing thing about this “317 breadboard” is that it’s dated 4/24/84 and has the notation, “Better tempco (temperature coefficient).” So perhaps Pease worked on this after Dobkin left the company. Dennis Monticelli wrote me, “Don’t remember it. I think Bob reported to Tim Isbell at this time. He reported to me later in 80s. That said, it fits with Bob’s style. He was always exploring ways to improve bandgaps. He may have done this project on the side and, if successful, would have advocated for making a premium version. The 317 was being widely copied at the time so performance differentiation was probably on his mind.”
PCB Boxes For RF
Soon Grohe fished another treasure out of a box (Fig. 2). This looks to be an RF circuit, probably part of the AM radio system National Semiconductor worked on decades ago. This little cube is of interest to show how Pease or a technician would fabricate a prototype that was suitable for RF frequencies. It’s a shame that some of the sockets are empty.
2. Pease and other engineers would use copper-clad circuit board material as a structural and shielding component in their prototypes.
There were two copper-clad boards for the LM4096. The reference section was dated 4/18/91 and initialed by Pease with a notation of his tech, JC Huerta (Fig. 3). The board had over a dozen transistors in metal cans. These are “kit parts” that design engineers could use to make a physical prototype of an integrated circuit. The kit parts were made on the same process as the IC. The kit parts will have more inductance in the wiring, and they won’t track temperature as well as when on the same IC. Even with these limitations, it was a great way to see if a circuit would work, before Spice simulations became so common.
3. This is a prototype of the reference section of the LM4096 1-W audio amplifier, no longer in production.
The circuit board connected to the reference section had more kit part transistors and a precision resistor IC with eight 10-kΩ resistors (Fig. 4). Being in the same IC, these resistors do track over temperature. This board uses a prototype method I had not seen before. The board has a matrix of holes that fit the metal can of the kit parts. Some of the cans are soldered to the copper-clad circuit board. This is similar to a “dead bug” prototype, since the parts are upside down, and resemble a dead insect. With the kit parts firmly mounted, it’s simple to “air-ball” wire the leads and connect in discrete components.
4. A second board Pease used in the LM4096 prototype.
I looked in a large box that was formed by two boxes taped together (Fig. 5). It was Pease’s ball-on-beam demonstration, where he showed that an operational amplifier can balance a ball on a moving beam. Pease used this demo at many trade shows and seminars.
5. A large box contained Pease’s ball-on-beam demonstration.
There were moving boxes chock-full of prototypes, circuits, and other Pease detritus (Fig. 6). One box had a very complicated copper-clad prototype (Fig. 7). I only wish that Pease had kept better documentation on them. When I build prototypes, I have a ring binder for them. Each prototype of a set gets a section where I keep track of any changes or differences to the others, as well as the test results. The rest of binder is full of datasheets and bills-of-materials for the design. Rather than keep a lab notebook separately, I would just put my design notes in one section of the binder.
6. A moving box from Bob Pease’s home had several prototypes and a CD ROM player, as well as lots of paper. Pease was hoarder.
7. Another box from Pease’s home has a more complicated prototype, a switch panel, and dozens of odd and ends.
Yet another moving box had a cache of complex hand-built Pease prototypes (Fig. 8). Oh, if these slabs of copper-coated FR4 could talk. On the top was another prototyping technique I had not seen. Pease cut copper-clad board material into long narrow strips. He then soldered three of those strips to other boards, spaced out so he could solder components in between. I could see the strips carrying positive and negative power, as well as a ground strip in the middle. You could use the other sides of the copper-clad to interconnect intermediate circuit nodes. Pease was clever in everything he did.
8. This box has several of Bob Pease’s hand-built prototypes. The 9-V battery is most likely long dead.
There’s Gold in Those Boxes
One box was full of gold-plated metal-can parts (Fig. 9). I thought those were valuable, but Grohe explained that they would need to be scrapped. Since they might be what is called “first silicon,” there was no way of knowing if the parts were production parts or prototypes. The records of the case codes are long gone.
9. Pease had a huge hoard of parts in his lab. Some of these are unlabeled and may have been for his gag Czar of Bandgaps costume used an advertisements.
When I worked at National Semi, I once saw 4,000 LM-series audio boomer amplifiers in tubes in a hallway. Thankfully, Grohe had me check with the audio group, and they confirmed they were a first spin of a part, and had problems, which is why they were scrapped. They went to a metal recycling place, not an electronics salvage yard.
Not all of the boards were hand-built prototypes. Bullock found a TP3237V evaluation board (Fig. 10). This is not a TI board or something Pease designed. There’s a company name, but several letters are obscured by a wire. It looks to be something like Semoltec, but I can’t be sure. I suspect we threw this board away since it was not a Pease prototype. Likely, it was something Pease used to evaluate one of his parts or some other design experiment.
10. A mystery evaluation board. Searching the part number TP3237V does not yield any meaningful results.
A nice little prototype was labeled “B-B Mult” (Fig. 11). This means Burr-Brown multiplier, a circuit that multiplies two input signals. Pease loved any good analog IC, even if it was made by a competitor like Burr-Brown. How nice that both National Semiconductor and Burr-Brown were both bought by Texas Instruments.
11. Pease fashioned this voltage multiplier circuit using Burr-Brown integrated circuits.
Another cute prototype used a Linear Technology part (Fig. 12). It’s an LTC1657 DAC (digital-to-analog converter) IC. That explains the DIP switches Pease used to enter a digital value. We have no idea why this was built on clear plastic as opposed to copper-clad circuit-board material. It might have been an artistic statement, or something to show off to Pease’s friend, Jim Williams. Or perhaps it was a Jim Williams prototype that he gave to Pease.
12. A DAC circuit Pease fashioned on clear plastic, very unusual for him.
A VHS Cassette Box… a Hair Dryer… and a Thermos Mug
Pease was an avid videographer, carrying about his VHS recorder to treks in the Himalayas as well as trade shows. I once saw him in Fry’s Electronics taking a video of the long line. It was no surprise he built a prototype inside a VHS cassette box (Fig. 13). Once again, we are not sure of the function of this prototype, with two ICs and a handful of transistors and discrete components, it might be anything.
13. Pease used an old VHS cassette tape box to house one of his prototypes.
Pease did not design a lot of power electronics, so I suspect we found an audio amplifier prototype (Fig. 14). With what looks to be two power rails and two metal-can transistors, this is likely a stereo amplifier. I know Pease worked on the remarkable LME49710 audio operational amplifier, so maybe that’s what this was for. We were in a rush, so I did not jot down the part numbers of the ICs in the prototypes, sorry.
14. A more substantial Pease prototype, likely a stereo audio amplifier.
Some of the prototypes went back to Pease’s time at Philbrick Researches, in the 1970s (Fig. 15). This assembly incorporated a Philbrick PP65AHU operational amplifier module on a small circuit card. Another prototype had the mezzanine card removed (Fig. 16). There you can see how Pease would put heat shrink around three TO-92 transistors so that they would track over temperature. The back of the board showed how Pease hacked in a few discrete components (Fig. 17).
15. This mystery circuit card incorporates a Philbrick PP65AHU op amp.
16. Another prototype board. but it’s missing the mezzanine card.
17. The backside of one of the prototype Philbrick cards, showing some cuts, jumps, and added components.
18. Another audio amplifier prototype, with transformer-coupling on the output.
Some prototype functions will be lost to history (Fig. 19). This one was tangled up with a hair dryer that Pease surely used to test his circuits over temperature extremes. Be careful if you do the same, it’s easy to overheat an IC with a hair dryer.
19. Many of the boxes from Bob Pease’s house were jumbles of stuff.
Grohe was delighted to find a thermocouple assembly he used when he helped Pease design an ultra-stable and ultra-precise temperature bath (Fig. 20). He could regulate it to 0.1°C, and the entire device was built into a Thermos mug. They used this to test Pease’s LM35 temperature-sensing IC.
20. Paul Grohe shows the thermocouple assembly he helped Pease build.
We also found an LM109 test fixture (Fig. 21). It was built by analog aficionado Fran Hoffart, now retired. Dated 2-16-75, there’s a label on the right side that says “Widlar’s Butt.” I thought this was a crude reference to Bob Widlar’s anatomy, but it stands for “Button.”
21. Not a Pease prototype, but a test fixture built in 1975 for Bob Widlar’s LM109 voltage regulator. The “Widlar’s Button” on the right is a small hole in the board that will activate an optosensor when you put your finger over it.
I got an explanation from Bob Dobkin, now of Analog Devices. He wrote, “The test box, which has an object on the right side called Widlar’s button, was developed around the LM309. Widlar didn’t like the box at first and wanted the push button on the right side, not the left side. So we took and scraped the copper clad off the bottom inside of the box and put a photo detector underneath that area. When you put your finger over the Widlar button, it blocked the light and activated the test sequence. It was always fun playing tricks on Widlar. When he put his finger on the Widlar button and it activated, he almost fainted.”