Post subject: transient freq. vs. ic and transistor sizing
Posted: Fri Dec 30, 2005 12:51 pm
hello...i am designing a RF
amplifier for a certain gain. the transistor i am using provides
the usual plot for ft vs. Ic for a certain dimension...but i need
to operate at a Ic value which is higher than the Ic that gives
max. ft....how do i insure that i can use the Ic i want and still
have that Ic correspond to max. ft....obviously i increased the
transistor size but that would drop my ft at the Ic i want.
the related question how can i find out where the max. ft is
for my modified transistor size...is there any way of plotting ft.
vs. ic by yourself using cadence/ads, etc or in other words given
a certain transistor size and Ic value how can i figure out what
the ft is for size/Ic combination...thanx much
Post subject: FtPosted: Fri Dec 30,
2005 1:35 pm
The issue isn't what the transition frequency ft
is - that's the frequency where your transistor becomes as useless
as a lump of coal. The issue is what gain (hfe or beta) you have
at the operating frequency. People often use the "rule of thumb"
that the amplifier should run at no more than ft/10, but that's
not always possible. I encourage you to look beyond ft.
obviously designing an IC - the ordinary designer, with discrete
devices, has no control over the geometry. Unfortunately, the ft
versus ic curve is characteristic of the process and layout specifics
of the transistor - so you don't really have much flexibility here.
This is really an issue you should take up with your vendor's
support staff. If they can't help you, you might want to re-consider
working with them, as they would be either uninterested in your
order (a bad sign), or incapable of intelligently analyzing their
product (also a bad sign).
Post subject: Posted: Fri Dec 30, 2005 2:00 pm
hello...but isn't there a way to estimate what my ft is using some
sort of analysis?? let's say i have fixed a size and Ic...now just
want to know what my ft would be....i remember something about you
can do ac analysis and then something but i don't recall it
Post subject: Posted: Wed Jan 04,
2006 11:14 pm
Joined: Sat Dec 31, 2005
Making the transistor bigger does not necessarily
cause the peak ft versus Ic to drop in frequency.
you do, DO NOT bias the transistor beyond the peak ft Ic because
ft drops rapidly after the peak and the ft will vary a lot part
to part for the same bias current.
You can always test ft
in simulation by running parametric simulations sweeping both the
frequency and the base input current. Then you need a function that
finds the peak ft frequency for each value of collector current.
Then you can plot peak ft versus current density or just current
if you like.