HP Memories

Return to Index

Some Memories of a Hewlett Packard Product Designer, by Bob DeVries

- Part 2 -

Test Department 1956-59

Part 2 - Table of Contents:

Test Dept 1956-59

First day at HP with more to follow

On April 16 th , I went to HP and asked for Don Peebles. He introduced me to Jim Hobson who was a line leader in the test department. Jim was a great guy and my boss for several months. The test department was fun and interesting. I didn't realize all the instruments being built and tested. Lots to learn here. After a tour of the facilities we got down to business and I was introduce to the 212A Pulse Generator. I was shown how to test them, what to look for, adjustments to make, test card to fill out, and where to put the tested ones. I thought I had maybe four or five to test but production runs were usually 25, 50, or 100 instruments. Maybe we had 25 of the 212A to test. Actually they all tested pretty well. I got through my first run OK with no rejects.

HP just finished designing the 211A Square Wave Generator. I was assigned to test this first or second run of 100 units. As new instruments were tested one would check operation at high and low line, calibrate the output to match dial readings, make sure it has the proper output, etc. Well, The 211A has a Symmetry control that changes the frequency output depending on the setting of that control. I had about 25 units rejected because I didn't know the Symmetry affected frequency. I was told that the symmetry of the square wave should be set for a symmetrical output before the frequency was calibrated. I had to re-test and re-calibrate about 25 Square Wave generators. That wasn't too bad and about the biggest mistake I ever made in the test department.

Recalling many early details

One must realize that recalling happenings of fifty years ago is quite a challenge. That is especially true for a slow learner/fast forgetter like me. This section will cover some of my home and work activities and learning experiences, as a "Test Engineer" in the test department at HP.

My test department activities covers a little over three years. When I joined HP, I think they had about 600 employees. I was quite sure I'd be working in test for my working career. I had no idea what the future would bring. I was quite naive and never worked for a large company like this before. I didn't know the history of HP and its potential for growth. I was just very happy to be off the road, living in California and have this fine new job working in electronics again.

A place for home projects

When we moved to Palo Alto we purchased a home at 3417 Cowper St. Our new home was almost within walking distance of my new job. I walked it a few times and biked many times. We moved to Palo Alto with our 14 months old daughter, JoDee. Our new home was more of a cracker box and built by Stern and Price builders. It had a wall furnace, two small bedrooms, a bathroom, kitchen, dining/family room and living room. It had no garage but a small shed attached to the front of the home and at the end of a short driveway. I had a lot of hand tools, power plane, router, drill press and table saw that I brought from Spokane. That shed worked quite well for my tool storage and workshop. This shop was about 10' wide and 6' deep and was a little crowded but not too bad. Any table saw work was done out on the driveway. Within a week I was putting all those tools to good use.

Reporting for work at HP

Jumping small hurdles and learning new things

I don't remember why, but one day, shortly after I was hired, I needed to make a copy of a drawing. I learned that in the office area there was a room that had all the drawing masters filed in logical order. There was a Bruning copy machine against the wall. Someone helped me find the drawing master I needed but I had no idea how to run the Bruning blue-line machine. There were lots of knobs and switches on the machine. Then there was the print paper that had one good side. Then what goes on top, the vellum or the print paper? Is the good side up or down? What print speed? I didn't know and I just didn't want to screw things up. So I asked one of the nice ladies near by and she showed me the copy procedure. Since then, I'm sure I've made a million blue-line copies of drawings but that first one was a memorable experience that scared me. I was so green and timid at first. I just didn't want to do something stupid.

General routine for a test guy

My general routine as a test guy went something like this. The new instruments were assembled and wired then placed on a four-shelved dolly. The dolly could hold about 16 or so large instruments waiting to be tested. There was a pre-printed test card accompanying each instrument with the instrument serial number noted on the card. The card acted as a guide and reminder to the test tech, to set and calibrate all the instrument's parameters for long-term reliable service. After the dolly was loaded it was rolled near my workstation. These instruments have just been assembled but had never been turned on. I lifted one off the dolly and set it at my workstation.

For a test tech, this is the exciting part. I'd look the product over for any obvious wiring errors or assembly problems. Are all the tubes in the correct location? If that all looked good, I'd plug it into a Variable voltage AC source called a Variac. I'd slowly bring up the AC to the product while monitoring the DC voltages in the instrument. I'd be on the alert for anything getting too hot or funny smells. If all that went well, I'd continue on with the test and calibration procedure. As the various parameters were tested the results were noted on the test card.

It didn't take long to become familiar with the new instrument, especially after testing several dozen of them. Sometimes an error was made in the wiring of the product. Maybe it would be a consistent error and the wiring/assembly line leader was called over to check for the error in other instruments waiting to be tested. Many times the wiring ladies would be asked to make the repair because their soldering technique was more professional looking than the technicians.

Sometimes wiring errors caused Allan-Bradley carbon resistors to overheat and they had a distinctive smell when they burned. So finding a wiring error or problem just amounted to following your nose and checking that part of the circuit that has the black or charred Allen-Bradley resistor. Sometimes the plate of a vacuum tube would glow red because it was overloaded due to a wrong or defective component or wiring error. That red glow caused the tech to focus on that part of the circuit for wiring errors or component problems.

Back then, there were no transistors and few semiconductor diodes in our products. Nearly all our products used vacuum tubes. Tubes run quite hot so many instruments required fans to keep the unit cool and air filters to keep out the dust. When changing tubes to solve an operation problem one could easily burn fingertips. All the tubes required high voltages to operate. Usually 200-400 Volts DC. When trouble- shooting for problems the tech had to be careful to respect those high voltages. It really shakes one up to get across 400 Volts DC.

There were some occasions where a couple of wires in a cable harness would short and that would cause the wires to get hot which would cause the plastic insulation on the wires to melt resulting in a ruined wiring harness. Some harnesses had more than 100 wires (200 ends) that had to be carefully disconnected. Then the new harness is re-installed and re-soldered in place. It was a very big job and a costly wiring problem to correct.

Most of the HP products of that time were point-to-point wired. One of the early HP instruments using the new technology of printed or etched circuit boards was the 523A Electronic Counter. And later HP came out with the 150A and 130A Oscilloscope. Otherwise all other products were interconnected the traditional way using tube sockets, cable harnesses, jumper wires, tie points and terminal boards.

A new baby, a new home and home projects

Much to my surprise about June of '56 Donna tells me she is expecting another little baby. Oh, my. . . that is good news! Now what'll we do? About July I got my first raise of 10 cents/hr. Boy, that'll help a lot!

While walking one evening, we passed some new tract homes being constructed in our neighborhood. They were called Eichler homes. They are of modern design, single level, slab floor with a lot of windows. We both fell in love with them and considered purchasing one if we could afford it. We realized that our present home would be a little too small with the new baby coming. A few days later, while on a drive in Palo Alto we passed another tract of new Eichler homes on Edgewood Drive. We stopped in at the sales office, talked to the Eichler folks, got a home tour and found the perfect home for us at 2047 Edgewood. It was a single level, fenced yard, three bedrooms, two bath, family room/dining room, Kitchen, two car garage and radiant heating. All this for just $18,950.00. On 9/15/56 we sold our 3417 Cowper home for a little more than we paid for it and purchased this new Eichler. A few days later I got another ten cents/hr raise from HP. Things are really looking good.

New 130A Oscilloscope with problems

The test stations are as follows:

1. 1- Turn on the scopes and set power supplies. Then heat-run for 24 hrs.
2. 2- Test and calibrate the Vertical Amplifier
3. 3- Test and calibrate the Horizontal Amplifier
4. 4- Test and calibrate the Sweep Circuits
5. 5- Do an overall test and check amplifier stability. Run a one-hour amplifier drift check on both vertical and horizontal amplifiers on the most sensitive range.
6. 6- Do a mechanical inspection, approve and send to shipping
7. 7- Do a spot mechanical and electrical check of scopes picked at random
8. 8- If there were no problems, the instrument would be packaged and shipped to a waiting customer

Item five above has an interesting side note. In order to accomplish this drift check I had two Varian Strip-chart Recorders. For the drift test, they were connected to the vertical and horizontal deflection plates of the oscilloscope under test. The scope sensitivity switches were set to the most sensitive range of one millivolt/cm. The spot was centered on the scope screen. The scope was allowed to warm up for 10 minutes with the cabinet on. After the warm up the spot couldn't drift off center more than a couple centimeters in an hour. Before the drift test I would smack the scope on the sides and lift the front and let it drop on the bench to check for microphonic input tubes. The paper chart recorded the spot drift in both the horizontal and vertical direction. If it drifted, tubes had to be changed and the scope re-calibrated. I would always pray for minimum drift.

Meeting Mr. William (Bill) Hewlett

I was still a new kid on the block and one morning this engineer looking guy (along with another fellow) came to my workstation and said he wanted to borrow one of my Varian recorders. He acted like he owned the place. Coincidentally, at that particular time, I had no instruments being drifted. The recorders were not being used nor connected to anything. I said "No, I'm sorry, but I use these all day every day!" I couldn't tell the man that it's in use but I did tell him that I use both recorders to run these important130A drift tests. "Can't ship the scope without the drift test", I said. Well, I looked around and saw that most of my co-workers mouths just dropped. The other man standing there couldn't believe what I said. I couldn't figure why all these folks looked so dumbfounded. Then I said to the short stocky engineer looking fellow, "If you really need one and my boss, Harvey Kellogg says it's OK, I'll let you have one." Everyone looked at me like I'd lost it. Then the engineer looking guy said, "Oh that's OK, I think I'll be able to find one somewhere else." Then both fellows left. Well, in just a couple microseconds my co-workers told me that I turned down Bill Hewlett and wouldn't let him use his own recorder. They couldn't believe I'd do that.

Actually that was the first time I met or saw Bill. Before this encounter I didn't know what he looked like. I didn't get fired nor promoted for turning him down. I just continued doing my final electrical evaluations and drift tests. It was an experience I'd never forget nor would my co-workers. Months later at a Christmas party, I told Bill that I was very sorry I'd turned him down. He was very gracious and told me that if I'm using something he needs, and I turn him down, that is just fine with him.

Raises are given and taken away and Rick is born 3/5/57

I loved working at HP for many reasons

Compared to a lot of jobs, I was just in heaven working in test at HP. All our instruments were clean and new. It wasn't like fixing old radios and TV sets. All the co-workers were friendly, helpful and smart. The environment was extremely stimulating. There were interesting things going on in adjacent HP buildings. Every chance I could, I would go for a walk around the plant. We had a casting shop, a machine shop, a plastic molding shop, a sheet metal fabrication shop, a paint and drying oven area, an engraving department, a machine assembly department, a tube aging area, and the various wiring and assembly lines. We had a fine library of technical books, a customer service department and a test equipment maintenance and calibration area for our own test equipment. We had the lab where new products were being invented, an environmental test area along with shake table and screen room for electromagnetic interference tests. We had a publication department for service manuals and advertisements. There were lots of folks in the offices processing orders, paying bills, doing bookkeeping and others working on mysterious things. As a new kid on the block I had a lot to learn, no doubt about it.

I haven't mentioned the beer busts that would occur more frequently than one would even desire. I attended a few but I really didn't care too much for lots of free beer and Donna and the kiddies were home waiting for me. They were always fun though with a lot of social interaction and camaraderie.

Useful instruments from rejected parts

I mentioned earlier the Friday auctions of junk and rejected parts that Don Peebles conducted. The HP quality standards were very high and many times parts were rejected because of a minor flaw. It wouldn't take long to collect all or most of the parts to build a fine HP instrument (with a few flaws) for home use. Missing or needed parts could be purchased from HP for cost plus ten percent. Sometimes the auctioned parts needed repair.

Early on I built a 211A Square Wave Generator and a 521A counter at very little cost. The fifty-cent decade dividers for the counter didn't work at first try but with a little trouble-shooting I was able to make them work just fine. How lucky could I be to own these expensive HP instruments for just a few dollars and a little work. The photo above is of my electronic work bench at 2047 Edgewood about 1959. Shown there are several instruments built from auctioned parts that I assembled and worked well. The photo includes a HP 150 and 130 oscilloscope, a sawtooth generator, a couple meters, a signal tracer (brought from home), a frequency/voltage standard and a 521 electronic counter. What fun to have such swell equipment.

As I became more familiar with the company and while on my lunchtime walks around the plant, I'd look in the trash cans for any useful items that were rejected for one reason or another. I'm a little embarrassed to write about it now but it happened. These parts were too poor to even be included in Don's auctions but with a little work could be made to work or be used for something entirely different. At the end of the week, I'd just ask Don for a property pass to take the junk home. I always had one electronic project or another going on at home in addition to important home activities.

The Harmony Plotters

HP had a social group called the Harmony Plotters (HP). There were representatives from each department. I was asked to represent the Test Department. Don's junk auctions brought funds into the Harmony Plotters treasury. The treasury was used to fund social events for the HP employees. Even the Harmony Plotters had their own annual auctions of obsolete instruments and hardware. More funds for the treasury. I was asked to be the auctioneer for one of their auctions. What fun that was. We sold lots of good stuff and I even purchased some good junk for myself.

Things were continuing to go along well at HP. During this period HP continued to grow, they leased some land at 1501 Page Mill Road on Stanford property and built several new buildings. To celebrate the completion of the first new building on Stanford land, the Harmony plotters proposed a lovely 1958 Christmas party on the upper floor of the new building. The floor was an acre in size. Lots of room for lots of people. I took an active part in food preparation and serving of drinks. We had live music, dancing and happy conversation. The party was a complete success and memorable in many ways.

Oscilloscopes, 130A, 130B, and the new 120A

After the big party, many assembly and test operations were transferred to the "Hill." Our line continued to build and test the 130A and 130B. Because of warranty problems using printed circuit boards, Dave Packard said that there would be no new products designed using PC Boards until we can make them more reliable. In the meantime, the lab was designing a new Oscilloscope called a 120A. It was released to production using point-to-point wiring.

It's interesting to note that Duane Dunwoodie was the project leader on the new 120A scope. Bill Jarvis was the marketing manager for the scope group of HP. It wasn't too long after the release to production of the 120A that Duane along with a couple other HP engineers (Bill Jarvis and Pete Lacy) left HP and started up a new company call Wiltron Electronics. Wiltron was purchased much later by Anritsu of Japan and became Anritsu Company in Morgan Hill CA .

My friend, Eric Hammerquist was the product designer on the 120A. Over a few months he and I became good friends. I would frequently go up to the lab and have lunch with Eric and discuss product design and other subjects of the day. I owe a lot that I learned about product design to Eric. A few years later he moved to Fluke Electronics in Seattle .

PC Board reliability problems and solutions

The 130A, 130B, 150A oscilloscopes and 523A counter were still being built and shipped using PC boards. But, there were warranty problems, especially on the 523A and 150A. Both had very large PC boards. HP tried two things to help solve their reliability problems. First, someone got the bright idea of using steel wool to clean the PC board surface before loading the components and dip soldering. That was a disaster. The steel wool cut through the copper traces with a microscopic fine line and the solder didn't jump the line so there was an open circuit in many places on some boards. Trouble shooting the open circuits was very difficult. As I said, the break (or cut) was microscopic so one couldn't see it with the naked eye. In addition, the steel wool left minute shreds of steel on the board surface and when the boards were dip-soldered, the circuits were shorted by the steel shreds making trouble shooting even more difficult.

Also to ensure vacuum tube reliability and to weed out the marginal ones, they put 100 tubes at a time on the shake table and would vibrate them for several minutes. That helped detect early failures in some but actually did more harm than good for long term reliability of the tubes.

The technique that really helped PC board reliability was to put an eyelet (hollow rivet) at each through-hole on the board. This would require several hundred eyelets per board. The eyelet made a good mechanical connection to the board traces and helped hold down the copper traces to the phenolic PC board substrate. Using thicker copper helped and using no steel wool for cleaning was the answer. Years later PC board technology improved and the component holes were plated through which eliminated the need for eyelets

Semi-automated testing of oscilloscopes

While testing these oscilloscopes one used several required instruments at his test station. Throughout the test procedure it was necessary to connect various signals from these instruments to the input connectors of the scope. One was always moving the connector from one test device to the next. I thought of a simple test fixture to connect all the cables from the various instruments to a switch box then used only two cables from the switch-box to the scope inputs. Now the technician just rotates a switch to select the various required signal sources to calibrate and test our oscilloscopes. This proved to be very helpful in speeding up and simplifying the test procedure.

Redesigning the 120A for better profit

We accepted the 120A oscilloscope into production and built and shipped many of them. The only problem was that this oscilloscope was labor intensive because it had tube sockets mounted on a fairly deep and narrow aluminum chassis and used jumper wires, tie points and terminal strips to mount and interconnect all the components. Trouble shooting and repair was difficult because the many components were hard to access. It was good looking and well made but it was expensive to build and repair. As test line leader I figured there must be a better way. I attended a meeting with the appropriate production engineers, production managers, engineering folks and even Dave Packard. I presented my observations as the test line leader, the things that made testing this product difficult. We came up with some small items to implement to make assembly and test easier but I was convinced that bigger improvements could be made.

In early '59 I was talking to Bill Bohnett, my friend in the PC fabrication department. During the past few months, PC fabrication and PC reliability had improved a great deal. Bill and I agreed that we could redesign the 120A put it on PC boards and really lower the assembly & test costs. We did it "under the counter" without management's permission or knowledge. Looking back, that might have been a dumb Idea. I'm sure management would like to know when and where they were spending their resources.

An instrument redesign is a big project for any individual and usually would be done in the lab. But for us to sell the redesign idea to the lab would be even more difficult than just doing the redesign ourselves. There were no circuit design changes but only mechanical design changes required. Bill and his department had all the facilities to fabricate any PC boards that we required. He had access to the dark room for photographing tape masters. He could make silkscreens to print circuit layouts on copper PC laminate. He could etch away the unwanted copper and drill and eyelet the through holes. He could load and dip-solder the PC boards. Bill asked Betty Downs to help us create the tape masters of each of the five circuit boards. So Bill and I agreed to build a PC version of the 120A. We informally called it the 120B.

Since I was familiar with the performance testing of the120A, I would know if the 120B performed properly and met specifications. I wanted it to work as well or better than the 120A. This redesign job was a little more difficult than just re-designing one product. The 120A had two versions. One was a cabinet model and the other was a rack mount. So I really had two products to think about. Throughout this project our scope line continued building and shipping various oscilloscopes. But, for several nights and weekends I worked on the new PC board blanks, chassis design, and switch layouts. I worked with Betty on the board layouts. I worked with Bill on PC board fabrication, board shape and mounting details. I worked with the sheet metal shop on the new board supports, front panel, side gussets and rear panel parts.

I was somewhat a fish out of water because I didn't have a drawing board nor the design knowledge of our experienced product designers in the lab. Again, Eric Hammerquist helped me on some of the documentation details. Looking back, there was so much I didn't know that I'm surprised we were able to get it all together with me as the project leader. I was making informal free-hand drawings and submitting them to the model shop for chassis part fabrication.

Little by little, our chassis parts were being made, the PC boards were ready to mount and test, the front panels were engraved, painted and filled, the chassis parts went together well and all the components fit and mounted as intended. I stayed late one evening and interconnected all the individual board assemblies, applied power, turned on the new 120B and completed the test and calibration procedure. It really looked clean, easy to build and it worked to specification. Bill and I couldn't have been happier.

With much excitement and some informality we showed this new oscilloscope design to our supervisors and managers who were all quite impressed. Even Packard, Hewlett, Ed Porter, Ralph Lee and Stan Selby came down to see this new product. Management accepted this PC 120A scope (120B) as a project they wanted to pursue. About that same time I was promoted from "Shop" to "Salary" on the pay system. That means I will be paid by the month rather than punch a time clock and be paid by the hour.

Transferred to the lab to finish 120B design

In a few days, I was told that I would be transferred to the lab to do the final design and formal documentation of this new 120B. I would be assigned to Norm Schrock's scope lab. I quickly learned that building a working prototype instrument is just the first step. In a new product there is lots more work to do. Create clean and readable documentation including a bill of material, assign part numbers, create material lists, create new test procedures, create new operating and service manuals. Design and build new production tooling for this new product. Build a pilot run of 10 rack mount and 10 cabinet mount versions, go through and pass all environmental tests. I had lots on my plate now. We asked one of my test line technicians, Bill Miller, to take over as scope test line leader.

Actually my transfer to the lab was supposed to be a temporary assignment but lasted thirty-one years. This was a big feather in my cap. Here I am, a young guy who had some basic electronics training in high school. I could fix most any AM tube-type home or car radio. I had four years of radar repair in the Air Force, I sold service stations supplies for my dad for three years and now I'm working in the lab of one of the finest electronics companies in the world. I'm helping to invent and develop fine new products. I've no degree. I've attended no college and I'm working elbow to elbow with many brilliant cream-of-the-crop degreed engineers of one type or another. I couldn't be in a better learning environment. Early on I made up my mind to keep my ears and eyes open and work as smart and hard as I could to do as well or better than those around me. This focused work ethic paid off in many ways over the years.

Throughout these three years Donna and I made numerous trips to Santa Cruz to bring the kiddies to visit with Donna's mom and dad. They would visit us from time to time also. Our yard continued to grow, I mowed the lawn and washed the car as required, life went on and every day was a new experience. There are lots of details that I've skimmed over. I hope you haven't found this too boring. That's pretty much what I did as a test guy at HP. After these few short years in test, I was really fortunate to be selected to become an engineer/product designer in the lab. My design experience and contributions in Norm's scope group continued on for about five years and is somewhat interesting. So the next segment will cover those activities.

Some Memories of a Hewlett Packard Product Designer, by Bob DeVries

- End of Part 2 -

Choose Next Chapter in Selector Below

Various Documents Linked from this Page for References: