You've shut 90% from the readers of this forum with these questions.
Maybe post a few phemts device names and applications if you really
expect answers...
He wanted a research project, and anything's better than
yet-another-robot. You can google this stuff, too, if it sounds
interesting. The trapping-dispersion thing in GaAs fets is well
known... transconductance changes at low frequencies, 10's of KHz
sorta stuff, because GaAs is full of defects that trap charge in
little niches in the gate region, sorta erratically. That also causes
DC offsets to change with bias/temperature history, sort of like the
first, dirty-oxide CMOS fets and opamps. I can imagine all sorts of
fairly easy experiments to play with trapped charges.
I haven't found any references to the possibility of gaasfets having a
bioplar transistor mode; I think I've seen it, and I guess it's
possible, but I don't have the time or a reason to explore it just
now. If some student was interested in this as a project, I'd help.
The other fun project I'd like to do some day is resolve
single-electron leakage steps in a cmos device. It might be possible.
Or use an eprom or a dram as an imager; that's been done, but it still
could be interesting.
Or, well, he *could* build a robot.
John
