aGaiNST MYSeLF
- - + zenov
- - + zenov_yzr@yahoo.com
More About Me
My Signs
. ..Dites Moi
RetinaL
- - + ZPS Cooking Session 2
- - + Room Warming FAT
- - + Room Warming ZPS
- - + KM Farewell Dinner
- - + KF Farewell Dinner
- - + LB CNY 2007
- - + World in Pixelations
- - + CAAS Days
- - + Renovation
- - + CSC Day
- - + CSC Day Eve
- - + Ma'at Costume Steamboat Party
- - + VAT 2007
- - + ROMPers KTV
- - + CSC Day OC Misc Pix
- - + Ma'at Sentosa Outing
+s.jpg)
- - + Korea Trip
+The+Banner+s.jpg)
- - + CSC Day '07 Banner Painting
+s.jpg)
- - + HJC X'Mas Eve
+Staybackers+s.jpg)
- - + CSC Day Preps
- - + ROMP-ing
- - + ZPS Miramar Outing
- - + 23rd B'day
- - + CSC MAF
- - + LB MAF
- - + ZPS MAF
- - + MAF 1101
- - + Ma'at Cyclohunt
- - + Tiong Bahru Market Supper
- - + Crystal Jade Outing
- - + Bintan Agro Holiday Pt 1
- - + Bintan Agro Holiday Pt 2
- - + Ma'at SOKA SD Outing
RUnnInG IMAGES
- - + Mich Do Event Dance
- - + CSC Day Dance Practice
- - + ZPS Vidz ~ 魅力燃烧
- - + Ma'at Chalet by tC
- - + Canon by funtwo
- - + bintan agro pool
- - + Ma'at Chalet: Attack on Junli
- - + MAF1101: Stuffing the mooncakes
- - + MAF1101: Shaping the mooncakes
- - + E3 Day 5: Talentime
- - + sammi mv: Hui Lai Wo Shen Bian
- - + sammi mv: Wonder Woman
- - + sammi mv: Arigato
__BloGWorK_::
The Other Side of The Fence
zenovia! Ancient Archives
zenov in NUS
zenov's Lyrics
zenovogy
The Inside of The Fence
zenov's Hatelist
zenov's Photo Cache
zenovia! in xanga
zenov Multiplies...
Resolutions Done ..
-+ Read 2 Chinese novels +-
-+ 台北人 +-
-+ Sling Bag +-
Resolutions Buanged
-+ Stockbroker's License (adjourned to 2008)+-
-+ Learn Piano +-
-+ Join an additional club after attachment +-
-+ Overcome my strangers anxiety (too diff to access)+-
-+ Master how to interact with people better (same as above)+-
-+ Kodak Z740 Zoom +- (no time to raise the $$$)
-+ Shed 2 kg by CNY +-
-+ Shed 5 kg by end of ICT +-
-+ IPPT Silver by end Oct +-
CLick! CLick!!
Fotos
Yahoo! Mail
NUS home
Discovery
Sammi S'pore Forum
Sammi M'sia Forum
Garfield
UB
Happy Tree Friends
Sogou
Gamespot
Sunday, February 3, 2008
Deciphering
Realized I was a bit cryptic in my previous post. And I am always not happy at being cryptic. lol
Anyways, background of my FYP:
I am supposed to come up with a set of codes (called a model) using a program called HSPICE, to model after a new device called TFET. Since TFET is an 'upgraded' version of MOSFET (which is the common device now), I am supposed to create a sub-circuit using existing MOSFET models that will have the characteristics of TFET.
In order to do the 'fitting', I have to go through a list of 140 plus parameters of MOSFET, to see which ones, when I changed their values, will result in the MOSFET acting like TFET.
For each parameter, I will at least use 3 values. Eg, for Value N, I will test 100N, 0.01N and -N. If in some cases the values caused the program to abort, I will use like '1', '-1', '0.01', '-0.01', 10000N, 0.0001N. But on average, a parameter will be tested for 5 values.
Some parameters are inter-dependant. Like RDSW can be used alone, but RDSWG will by default take on the value of RDSW. In order to test RDSWG, I have to do all sorts of combination that will result in RDSW and RDSWG not having equal values. That combination, will on average require 25 changes.
Throughout the school holidays, I found that TOX, DTEMP, VTH, NSUB and U0 results in the final curve looking like the one I needed. Now I only needed to lower the minimum value and change the slope at certain parts of the curve.
3 weeks back, I showed my Prof my codes, he said go ahead.
3 weeks of work later, I show him my codes, he said parameters L and W should be kept constant constant and then I should test the rest of the parameters.
3 days later, after enduring 2 night of fever to complete 3 weeks of work, I showed him my work again and he said for the device we are working on, there should be a WHOLE SET of parameters that should be kept constant instead of L and W (and the other parameters I painstakingly found) that I used.
He criticized me for including the other parameters (mentioned above) as constant parameters.
For eg, NSUB has default value of 6E16, and in my code, I reduced it to 1 cos it resulted in the more desirable curve. The other prof (my examiner, who also happened to like to pass comments nonchalantly when he had no prior knowledge to the matter and also happened to come from the very powderful country of India) said:
You should have used only between 6E14 and 6E18. You should not just simply sub in a value of 1 and just forget about it. No wonder you couldn't change the slope of the curve. U0 default value is 670, you shouldn't have used 22000. You should have known all these.
He irresponsibly accused me of:
1) Not doing things systematically
2) Not taking pride in my work
3) Not having an understanding of the importance of each parameter.
But the FUCKING PROBLEM is, 1) 2) and 3) are WRONG!
I tried 6E14~6E18 and there was no change in the curve. When I changed NSUB to 1, there was a FUCKING CHANGE towards the ideal curve I want. THAT WAS THE REASON WHY NSUB=1!
This is the same as U0.
I really don't know what those 2 guys are thinking.
I really don't know.
It's either they are too intelligent for my primitive brain to follow, or I have been too affected by the secular (read: contemporary) world to understand their Nirvana state of minds.
After CA2, I tried to focus on the problem, the task at hand.
I may not be capable of dealing with research work like the rest of the 'elites' in engin, but I am trying my best. I am doing my best.
Who likes to work on such jobs, enduring 5 days of fever with minimum food and money at hand only to be chided as not being serious in work?
AND I HAVE FIVE OTHER MODULES TO WORK ON.
I am so tired of the attitude by my profs. I can't move on when they are not ready to do so.
Realized I was a bit cryptic in my previous post. And I am always not happy at being cryptic. lol
Anyways, background of my FYP:
I am supposed to come up with a set of codes (called a model) using a program called HSPICE, to model after a new device called TFET. Since TFET is an 'upgraded' version of MOSFET (which is the common device now), I am supposed to create a sub-circuit using existing MOSFET models that will have the characteristics of TFET.
In order to do the 'fitting', I have to go through a list of 140 plus parameters of MOSFET, to see which ones, when I changed their values, will result in the MOSFET acting like TFET.
For each parameter, I will at least use 3 values. Eg, for Value N, I will test 100N, 0.01N and -N. If in some cases the values caused the program to abort, I will use like '1', '-1', '0.01', '-0.01', 10000N, 0.0001N. But on average, a parameter will be tested for 5 values.
Some parameters are inter-dependant. Like RDSW can be used alone, but RDSWG will by default take on the value of RDSW. In order to test RDSWG, I have to do all sorts of combination that will result in RDSW and RDSWG not having equal values. That combination, will on average require 25 changes.
Throughout the school holidays, I found that TOX, DTEMP, VTH, NSUB and U0 results in the final curve looking like the one I needed. Now I only needed to lower the minimum value and change the slope at certain parts of the curve.
3 weeks back, I showed my Prof my codes, he said go ahead.
3 weeks of work later, I show him my codes, he said parameters L and W should be kept constant constant and then I should test the rest of the parameters.
3 days later, after enduring 2 night of fever to complete 3 weeks of work, I showed him my work again and he said for the device we are working on, there should be a WHOLE SET of parameters that should be kept constant instead of L and W (and the other parameters I painstakingly found) that I used.
He criticized me for including the other parameters (mentioned above) as constant parameters.
For eg, NSUB has default value of 6E16, and in my code, I reduced it to 1 cos it resulted in the more desirable curve. The other prof (my examiner, who also happened to like to pass comments nonchalantly when he had no prior knowledge to the matter and also happened to come from the very powderful country of India) said:
You should have used only between 6E14 and 6E18. You should not just simply sub in a value of 1 and just forget about it. No wonder you couldn't change the slope of the curve. U0 default value is 670, you shouldn't have used 22000. You should have known all these.
He irresponsibly accused me of:
1) Not doing things systematically
2) Not taking pride in my work
3) Not having an understanding of the importance of each parameter.
But the FUCKING PROBLEM is, 1) 2) and 3) are WRONG!
I tried 6E14~6E18 and there was no change in the curve. When I changed NSUB to 1, there was a FUCKING CHANGE towards the ideal curve I want. THAT WAS THE REASON WHY NSUB=1!
This is the same as U0.
I really don't know what those 2 guys are thinking.
I really don't know.
It's either they are too intelligent for my primitive brain to follow, or I have been too affected by the secular (read: contemporary) world to understand their Nirvana state of minds.
After CA2, I tried to focus on the problem, the task at hand.
I may not be capable of dealing with research work like the rest of the 'elites' in engin, but I am trying my best. I am doing my best.
Who likes to work on such jobs, enduring 5 days of fever with minimum food and money at hand only to be chided as not being serious in work?
AND I HAVE FIVE OTHER MODULES TO WORK ON.
I am so tired of the attitude by my profs. I can't move on when they are not ready to do so.
Comments:
Post a Comment
new acquisitionz
Wallet w Acc
I-Trigue 2300
Nike Shoes
This Is Mi
Ferrari Red Zen Stone Plus
True Style Never Shouts
Ferrari Red Gurita Mouse
New F Square Bag
New MP3 Pouch
New Thumbdrive & Chain
Birthday Bear
Orange Watch
New Glasses
Nike Sling Bag
Nike Shoe Bag
OP Misc Case
;Qu'est-ce Que J'ai Fait
outing w zps gang, ate at qiji & witnessed a bloody fight. then went to bowl at kim seng, bot 4D & had dessert at tbp mac's.
MEs.:||:.aMIs
yukiko a.k.a xie zi (shoe)
shuhui¡Öliu
fwed¡Öong
angie¡Öng
U.B
ethan hawke (faux pas)
PR
Spokesperson
S.A.M (Secratary.Archiver.Maid)
babeliciousme
Nua Queen
tinywingz
bryan the older
merv
shuhui¡Ötan
J1
LB On The Blog
6g
同居密友(Cohabitor)
surcose
丘羽
chicella
Little Miss Salty
ArchiVes .
September 2007
October 2007
November 2007
December 2007
January 2008
February 2008
September 2008