Check out the finalists of the Superhero stories that the SpringSoft Community submitted for the Q4 2009 Contest!
Story A - Archana, nSYS Design Systems:
What challenge needed to be overcome?
Once one of my customers sent me a dump file in wlf format, the file size was very large and I was finding it difficult to load in another waveform viewer. Also the look and feel of the viewer is not friendly.
How was the challenge solved?
Then, I used the converters like :
1) wlf2vcd
2) vcd2fsdb
The first converter changed the dump into vcd format and then, I converted it to fsdb using the second converter.
Also, VCD can be converted to FSDB using vfast .
Then, I could view and analyze the problem using Debussy tool.
What satisfaction did you receive by solving the challenge?
I felt like a winner after satisfying my customer request on time and efficiently.
Story B - Laurent, ST Microelectronics:
What challenge needed to be overcome?
Evaluate the quality of IPs purchased to 3rd-party.
How was the challenge solved?
Initially by using Certitude's Metric mode: it gave us objective metrics evaluating the overall quality level of the verification.
Then, switching to Verification Improvement mode, we were able to feedback our IP supplier with specific areas that were not well verified.
What satisfaction did you receive by solving the challenge?
Certitude gives us objective verification quality metrics. This is a great added value when discussing with 3rd party IP vendors.
Story C - Vagner Pires:
What challenge needed to be overcome?
It was a beautiful Friday in summer, I had arranged with my family that after the end of the day I will find them and travel to the beach. The day began around a simulation very important for completion of a project, when at around 4 pm, the simulation stopped.
How was the challenge solved?
I said desperately:"My God my weekend with the family is over because of this. I suddenly had a brilliant Idea, I used Verdi to find the bug and resolves it. Bingo! The bugs were found and resolved quickly and the product would be completed without problems.
What satisfaction did you receive by solving the challenge?
On the way to the beach I thought: "Thanks to the tools of SpringSoft I'm going to travel with my family."
Story D - VP Engineering, USA:
What challenge needed to be overcome?
We needed to pulse 4 635nm laser heads at more then 500Mhz to obtain modulation that was needed to provide therapeutic light for our next generation of technology. Our current level of technology used an ATmega32-16 to modulate at no more then 8Mhz. Further, the design would need to be a portable application that would run on batteries.
How was the challenge solved?
There were several technologies that were examined to overcome the monumental task of modulating Laser Light at those higher frequencies. These are explained as follows:
1. The first key was to build a High Speed FPGA that would be slaved to the micro-controller. It would receive its parameters via an SPI port. It was decided to use the Stratix III line of FPGAs with the 2nsec technology. The design ultimately would configure 4 32-bit counters that were controlled via 4 internal PLLs that would allow very accurate modulation frequencies to be produced by the design.
2. Because we used the Altera part we added two immediate engineering problems that had to be overcome to use this part. The first was the increased heat produced by this design. The case was modified to accommodate an increased heat sink for the FPGA and two fans were added to provide forced convection to remove the heat. The second problem was the addition 3 power planes and power needs of this beast. These will be discussed later.
3. The fans were to be powered from the 5V power plane and since each required 350ma , the 5V power plane was upgraded to a DC-DC converter using the National LM2735XMF. This part allowed us to change our internal Battery power plane from a dual Li-Ion at 8.4V Optimum to a lower 4.2V Optimum power plane with more efficient power conversion plus since our new FPGA used a larger amount of energy, we were able to upgrade our battery to a 3600maH 3.7V Li-Ion Battery.
4. Because of the additional needs of the FPGA to support a 2.5V, 3.3V, and 1.1V power plane, the changing of the Battery voltage to 4.2 was advantageous. The 1.1V was made possible by using the National LM3878-ADJ and adjusting to 1.1V and the On Semiconductor NCP3335AMN250R2G to obtain the 2.5V supply.
5. Since the power requirements for the 3.3V power were high around 450ma, it was decided to use the National LM2735XMF again in a SEPIC configuration to float between boost and regulate from the Battery to obtain the 3.3V power needed to run the FPGA and the new Laser Diode Drivers.
6. To drive the Lasers at the higher frequencies we decided to use the LVDS standard to drive from the FPGA to the Laser head. We then needed to determine what kind of laser driver would be able to support these frequencies and be compatible to the LVDS interface. We decided to use the Micrel part, the SY88212 Laser driver and interfaced it directly to the FPGA via the compatible LVDS driver from the Stratix III part.
7. Our last problem was how to calibrate each laser head in such a tight combined location to get reliable 5mw optical energy on each laser head. To accomplish this we used the Microbridge Rejustor component. This part allowed us to build a simple calibration jig where we measured the optical energy from each laser head and electrically adjusted the rejustor to a set value.
What satisfaction did you receive by solving the challenge?
With all the work done over the last 18 months to bring this new product to market, I am absolutely amazed that we made it this far. It has been a long and arduous effort to get here but I am already seeing the benefits in what our technology is doing for our patients. Symptoms and problems that the older system could not even attempt to get close to solving are now solved in ways that appear almost as magic. I have used the product and know that my health has also improved. It is so cool to know that after solving so many diverse problems that you can see the results in patients benefiting from your long hours of work.
Story E - Engineer, India:
What challenge needed to be overcome?
Two years back, I was working on verification of a client project based on RFID protocol. Before transmitting, the data was getting scrambled and received data was being descrambled. The design engineer from Client side was new to this project and scrambler code has already been implemented by someone. None from their side, knew which type of scrambler is implemented - Additive or Multiplicative. We both knew the scrambling polynomial (It was well documented). But no diagram was given. My valid data from test case after scrambling was not meeting the DUT requirements and was being discarded by DUT. None of us was aware of what is wrong. We tried verifying scrambler logic using different initial seed values and different input data. But we were not coming to conclude anything. So verification was getting delayed
How was the challenge solved?
I suggested to the design engineer to provide me the scrambled value for a set of data inputs with different seed values. I implemented both type of scramblers at my side and got the scrambler results for the same data and with same seed value. By comparing above results, I found that some of our results were matching for Multiplicative scrambler case, but some results was still not matching, which later on found to be a bug in design. By above procedure, we came to know the type of scrambler implemented in DUT. I provided the client with the scrambler diagram
What satisfaction did you receive by solving the challenge?
1. The verification work started proceeding, which was a big achievement for us.
2. We got client appreciation.
3. The scrambler polynomial, its type and diagram got well documented
Read the rest of the entries from the Q4 2009 contest
