Tekmos' Blog

Every part that Tekmos manufacturers, has to be tested.  The wafer manufacturing process will naturally produce a percentage of non-functional parts, and we must identify them and reject them.  Tekmos uses a combination of inside and outside test resources to test the parts.  Tekmos has experienced sizable growth over the last two years, and as a result, we were finding ourselves limited by our test capacity.  To address this, we decided to buy additional testers to augment the two testers we already had.

Purchasing a tester is a complicated decision.  You need to determine how many pins the tester will need.  Most of our devices have less than 100 pins, though some have 356 pins.  Since tester cost is generally a linear function of pin count, it is important to buy the right amount of pin capacity.  You must also consider the economic tradeoffs of tester speed and vector memory capacity.

After months of internal deliberations, we decided to purchase a new Griffin III tester from HiLevel.  This tester has 128 pins with a 64M vector depth and a 100 MHz clock rate.  This is a significant increase in our capability, since our older testers were limited to 40 MHz and 8M of vector depth.

While we were negotiating over the Griffin II tester, we found another HiLevel on the surplus market.  This one was a model ETS780 tester, with 256 pins and an 8M vector depth.  It is a much older tester, but was a valuable at a reasonable price, so we acquired it as well. 

Along with the testers, we have acquired a new problem.  Each tester has a different mechanical interface.  In the past, we have just built a custom interface for a specific part to connect to a specific tester.  However, this approach has caused capacity problems as our volume has increased.  We frequently find ourselves in a situation where we need two different parts that are only testable on a single tester.  This results in the situation where one tester is running two shifts, while an adjacent tester is standing unused. 

Last month Tekmos traveled to exhibit at another high temperature conference. This time it was the GEA Geothermal Energy Expo 2014 at the Oregon Convention Center in Portland, Oregon. The expo is billed as "the world's largest gathering of vendors providing support for geothermal resource exploration, characterization, development, production and management." Since the folks from Tekmos have spent most of their careers involved with microelectronics, it was fascinating to talk with the attendees and vendors who manufacture and use the massive equipment associated with the geothermal industry. Even though the exploration for and harnessing of geothermal energy makes use of equipment that is thousands of times bigger than our microelectronic circuitry, many pieces of equipment cannot function without the much smaller electronic components like those that Tekmos designs and manufactures.

The expo gave us a chance to talk in depth with various equipment vendors about the performance and electrical characteristics they need most when choosing components for their new products. We were again reminded of the need to listen carefully and to clarify terms. Some of the vendors at the Expo talked about their requirement for maximum operating temperatures of 300 degrees without mentioning they were speaking in degrees Fahrenheit. Others at the conference needed 250 degrees operating temperature, without mentioning they were speaking in degrees Celsius, the units generally used in the electronics industry. The 300 degrees Fahrenheit, about 150 degrees Celsius, is pretty straight forward for Tekmos using standard bulk silicon technology but 250 degrees Celsius will require our SOI processing.

This second high temperature conference of the year for Tekmos yielded a number of new contacts as well as a chance to visit with other customers that we have known for some time. We also received a number of inquiries about our standard products where the high temperatures are not required. There were even some inquiries about Tekmos supplying parts that are going obsolete by the original vendors. While this has been a major focus of Tekmos throughout its 17 year history, we were a little surprised since our focus at this geothermal conference was "Hot Chips", our relatively recent thrust into the high temperature market.

As we often do, we also took some time to visit a local attraction. We had a tasty lunch at a great Chinese restaurant suggested to us by a local who thought we looked lost as we got off the light rail. We were primed for our visit to Portland's Lan Su Chinese Garden, the most authentic Suzhou-style Chinese garden outside China. Its name is poetically interpreted as "Garden of Awakening Orchids." It was assembled by 65 Chinese artisans who spent 10 months assembling and completing structures that were crafted in China. From the Hall of Brocade Clouds, past the Scholars Courtyard, to the Tower of Cosmic Reflections, the Garden is a sea of tranquility near center of the bustling city of Portland. It was a welcome period of peace and quiet between our travel to Portland and our manning the booth at the Geothermal Expo.

When it comes to reasons for Tekmos Team Members to gather for good food and celebration, they can always be invented, but this event was for our nation’s birth, the 4th of July, and a 3 day weekend. Every holiday Tekmos staff have a “potluck dinner” where each brings a favorite dish or, if one is culinary challenged, bring chips, corn on the cob or a store bought banana cream cake. Everyone is expected to pitch in.

On the menu for this 4th of July BBQ were hamburgers, bratwurst, corn on the cob, homemade potato salad by Alison, deviled eggs by Lynn, and celebratory mozzarella and cherry tomato hors d’oeuvre by Carole, hot artichoke spinach dip with grilled pita bread by Leslie, cole slaw by Zeta, and homemade apple pie by Jon’s son, Ryan.

Food specialties were shared from favorite family traditions and recipes. The designated Tekmos griller is Phil, Tekmos’ all around machine guru. Phil’s family owned a steak house so it was only natural that he would take Bison burger and hamburger to make specialty burgers. An added bonus is that Bison is a very lean meat and healthier. They were cooked to perfection.

The piece de resistance of the 4th of July Barbecue was Jon’s son, Ryan’s homemade apple pie, a recipe passed down from his great grandmother. The apple pies were served with fresh whipped cream, not from the container.

The 4th of July annual celebration of the founding of our nation is often celebrated with fireworks, picnics, barbecues, and favorite summertime dishes of hot dogs, hamburgers, potato salad and apple pie. You can’t get more American than that.

Business trips are usually intense.  You work hard, and are frequently up late in the evening.

To compensate for this, we try to add a little extra to each trip.  This month, four of us went to the HiTec conference in Albuquerque.  We had free time for dinners, and 5 hours at the end of the conference, between the show closing at noon, and having to be at the airport at 5.  That was plenty of additional time for something extra.

Albuquerque has a 10K foot mountain in the city limits, called Sandia Peak.  There is a restaurant located on top of it, which you access by a tram.  We decided to have our Wednesday dinner on top of the mountain.  It was a beautiful 20 minute tram ride up there.  Once there, we were unable to eat due to lack of reservations, so we went to the bar, and ate off of the bar menu.  I also had a conference call while up there, which I took while standing on a rock ledge, overlooking Albuquerque 5,000 feet below.  It was cold up there, and there was still snow on the ground.  But the views were incredible, and well worth the cold.  We stayed up there until dusk, and then took the tram back to the car.

After the conference ended, we went to see the Petroglyph National Monument.  This monument contains a number of Indian petroglyphs dating from 1300 AD that are carved into blocks of lava on the west side of Albuquerque.  We walked along several of the trails, checking out the petroglyphs.  There were hundreds of them to view. Being an amateur geologist, I was also impressed by the lava flow.  We were in a small canyon that at one time had been a hill.  The lava had flowed on either side of the hill.  In a mere 200,000 years, the hill had eroded, leaving the canyon we were in.

After checking that out, we toured the National Atomic Museum.  This is much better suited for engineering nerds, with its collection of rockets, old airplanes, bomb casings, and old radios.  I particularly enjoyed studying the engines of a Titan II rocket.  I had seen a detailed description of the SpaceX rocket engine, and so I knew what I was looking at.  It was also interesting to study the avionics and wiring harnesses used in the planes and rockets.

Last month, Tekmos presented a paper at the HiTec conference. HiTec is dedicated to high temperature electronics, ranging from the relatively cool 175C up to the 1200C range for Silicon Carbide.  It also covers high temperature materials and passive components.

The Tekmos paper investigated the use of refresh cycles to assist with data retention in EEPROMS.  We are developing a new 256K, high temperature, and high reliability EEPROM.  At elevated temperatures, the charge stored in the memory leaks away, and must be replenished.  While an EEPROM at 85C can be expected to hold the data for 10 years, at higher temperatures (~200C), the data retention time is measured in days.  Restoring the data is made more complicated since the number of times that an Erase-Write cycle can be used also decreases with temperature.  The 100,000 cycles guaranteed at lower temperatures becomes 100s of cycles at high temperature.

We get around this by using a refresh cycle.  A refresh cycle does not do an erase prior to the write, and so can only be used to re-write the same data.  We were concerned that repeated refresh cycles could also cause damage, but our experiments showed that a refresh cycle is safe.

To prove this, we set up a system using our TK68HC11E1 devices, which contain 512 bytes of EEPROM.  This is the same basic EEPROM cell that we will use in our 256K EEPROM.  We attached 77 of the microcontrollers to a board, and configured them so that we could communicate with a specific microcontroller.  Then we downloaded programs into the microcontrollers, and ran experiments.  The experiments included such things as determining the number of Erase-Write cycles it took to burn out an oxide, or just giving the part a fixed number of refresh cycles.  This made it easy to collect a large amount of high quality data that was the basis of our paper.

For the first time, Tekmos had an exhibit at HiTec.  We showed our TK68HC811E2 high temperature microcontroller, and our 250C ASICs.  We also provided a preview of our 256K EEPROM, our 250C TK80H51 microcontroller, and the TKH4003 FPGA.  These last three products are under development, and we will be introducing them over the next year.