Another hectic week started which didn't allow me to write my blog as well. I did some good work in last few days. On 30th Nov, I spend lot of time in study, trying to realize the method to compare two set of probabilistic waveforms, one from an individual inverter with an ideal ramp input signal and another one from an inverter chain. The main challenge of this experiment to make sure that comparison of waveforms are valid, I mean the input of the final inverter in the inverter chain and individual inverter in another experimental setup should have same input signal slew. I have developed a method to size the inverter chain on the fly to generate the exactly same slew as of individual inverter. After making the required setup, I found that both waveform are almost same and having very close output slew. My next task was to change the output signal slew to analyze the waveform variation. I attacked this problem by changing the input signal slew and captured all the waveforms. Surprisingly, I found that output waveform in all the cases are having almost same output signal slew. This new issue encouraged me to play with circuit and come-up with set of parameters and their range to cover required output slew range from 5ps to 600ps. I started with input signal slew and channel length. This combination did not worked out well. Then I tried with input signal slew and channel width. This combination also fails to give full swing in the output signal slew. Till now, I was using a high capacitor as a output effective capacitive load. I had reasoning for making this constraint in the design. But, since none of the combination of parameters were able to give us full swing of the output signal slew. I had thought to change output load as well. So, in the new experiment, I have used input signal slew and output capacitive load as two parameters and it works! I got full swing of the output signal slew which was desired but on the cost of variable output capacitive load. Now, in the rest of the week, I have to develop the complete method to model a set of probabilistic output waveform with circuit component.
This was my two days of academic work. But today is my third day of writing blog. It means one day is missing. Exactly! I was not in the university on Dec 1. On that day, I went for an excursion to Dutch Space, A dutch Company. This excursion was organized by student group of our department (ETV). This company makes solar panel for space satellites, parts of satellites rocket launcher, robotic cars like one went to Mars, robotic arm for international space station and many more. In fact we saw all these products in the lab. This was an amazing experience for me. I have taken some photos of the trip which are posted here:
Good Night
No comments:
Post a Comment