Maker Pro
Maker Pro

FETs Vesus Bipolars, Why More Efficient?

D

Dr. Slick

Jan 1, 1970
0
Hi,

I've been using Mosfets for quite some
time now, and have enjoyed the greater
efficiency and gain that you get with them,
versus BJTs in VHF radio frequency applications.

I know the greater efficiency has partyly to do
with the much higher low frequency impedance
of the gate versus the diode of the base-emitter
junction, but i'd like to know the other reasons
if someone could explain them to me.

Thanks for your input.

Slick
 
K

Ken Smith

Jan 1, 1970
0
Hi,

I've been using Mosfets for quite some
time now, and have enjoyed the greater
efficiency and gain that you get with them,
versus BJTs in VHF radio frequency applications.

I know the greater efficiency has partyly to do
with the much higher low frequency impedance
of the gate versus the diode of the base-emitter
junction, but i'd like to know the other reasons
if someone could explain them to me.

Thanks for your input.


Here's one:

The storage effects in the bipolar makes you class C amplifier have a
wider conduction angle. In a MOSFET, the turn on and turn off speeds are
much more equal so the conduction time is delayed but not increased.
 
A

Active8

Jan 1, 1970
0
Here's one:

The storage effects in the bipolar makes you class C amplifier have a
wider conduction angle. In a MOSFET, the turn on and turn off speeds are
much more equal so the conduction time is delayed but not increased.

Low Rds(on) for two - that and the aforementioned fast switching
work together to lower power dissipation... but *how much* does that
really help at VHF?
 
P

Paul Burridge

Jan 1, 1970
0
Hi,

I've been using Mosfets for quite some
time now, and have enjoyed the greater
efficiency and gain that you get with them,
versus BJTs in VHF radio frequency applications.

I know the greater efficiency has partyly to do
with the much higher low frequency impedance
of the gate versus the diode of the base-emitter
junction, but i'd like to know the other reasons
if someone could explain them to me.

I think you meant to say "...much higher *high* frequency
impedance..."?
 
K

Ken Smith

Jan 1, 1970
0
Dr. Slick said:
No. The gate is like a capacitor in a FET, so at
higher frequencies, the reactance of the cap is small,
and you only have the ohmic (real) losses.

Don't forget those inductances.
 
P

Paul Burridge

Jan 1, 1970
0
No. The gate is like a capacitor in a FET, so at
higher frequencies, the reactance of the cap is small,
and you only have the ohmic (real) losses. At lower
frequencies and DC, the capacitive reactance is very
high. Xc=1/(2Pi*F*C), where F is the frequency in
cycles/sec.

Mea culpa. I thought you were referring to the signal impedance across
the B/E diode in a BJT.
 
D

Dr. Slick

Jan 1, 1970
0
Paul Burridge said:
Mea culpa. I thought you were referring to the signal impedance across
the B/E diode in a BJT.


Well, i kind of am, in the sense that the base-emmiter
diode has significant current across it past the threshold
voltage, whereas the FETs have more of a capacitor for the
gate-source where very little DC current flows.

That's why a BJT is more like a current-controlled-
current-source, and the FETs are more like voltage-controlled
-current-sources.


S
 
D

Dr. Slick

Jan 1, 1970
0
Don't forget those inductances.
--

yeah, the series parasitic inductance too, but
you have to get into UHF or microwaves before
they get bad.


S.
 
K

Ken Smith

Jan 1, 1970
0
[email protected] (Ken Smith) wrote in message


yeah, the series parasitic inductance too, but
you have to get into UHF or microwaves before
they get bad.

Depends on what you call bad. With higher power MOSFETs (BLF276 IIRC) the
input impedance hits zero angle before 200MHz so at 100MHz, it will be
effecting the impedance.
 
K

Ken Smith

Jan 1, 1970
0
Well, i kind of am, in the sense that the base-emmiter
diode has significant current across it past the threshold
voltage, whereas the FETs have more of a capacitor for the
gate-source where very little DC current flows.

That's why a BJT is more like a current-controlled-
current-source, and the FETs are more like voltage-controlled
-current-sources.

At high frequencies, both look like charge controlled current sources.
The normal base current of a bipolar is a leak that takes the charge away.
 
P

Paul Burridge

Jan 1, 1970
0
Well, i kind of am, in the sense that the base-emmiter
diode has significant current across it past the threshold
voltage, whereas the FETs have more of a capacitor for the
gate-source where very little DC current flows.

That's why a BJT is more like a current-controlled-
current-source, and the FETs are more like voltage-controlled
-current-sources.

I can see where you're comin' from; but you're gonna upset a lot of
folks here with that last para!
 
D

Dr. Slick

Jan 1, 1970
0
Depends on what you call bad. With higher power MOSFETs (BLF276 IIRC) the
input impedance hits zero angle before 200MHz so at 100MHz, it will be
effecting the impedance.

Parasitic inductance is usually less than about 2nH,
so it would be about XL=2*pi*f*L

So the inductive reactance would be about 1.25 Ohms
at 100 MHz. Certainly affecting the impedance, but
still pretty low.


S.
 
K

Ken Smith

Jan 1, 1970
0
[email protected] (Ken Smith) wrote in message


Parasitic inductance is usually less than about 2nH,
so it would be about XL=2*pi*f*L

So the inductive reactance would be about 1.25 Ohms
at 100 MHz. Certainly affecting the impedance, but
still pretty low.

Instead of counting on my memory. I dug out the BLF276's data sheet.

At 100MHz Zin = 1.3 -j1.7 (points taken from graph)

A 1.25Ohms Xl would be very important in this case since it is about as
big as R and Xc. Th package on this is a 6 pins as source configuration.
 
D

Dr. Slick

Jan 1, 1970
0
Paul Burridge said:
I can see where you're comin' from; but you're gonna upset a lot of
folks here with that last para!


Why do you think that? CCCS for bipolars and VCCS for
FETs is standard in most texts.

That's why when you look at a I-V curves, for BJTs the
lines are denoted with base current, and for FETs the lines
are denoted with gate voltage.


Slick
 
W

Winfield Hill

Jan 1, 1970
0
Dr. Slick wrote...
Why do you think that? CCCS for bipolars and VCCS for
FETs is standard in most texts.

That's why when you look at a I-V curves, for BJTs the
lines are denoted with base current, and for FETs the lines
are denoted with gate voltage.

Nope.
 
P

Paul Burridge

Jan 1, 1970
0
Why do you think that? CCCS for bipolars and VCCS for
FETs is standard in most texts.

FETs are certainly accurately modelled as VCCS in a certain regions of
operation. BJTs are *inaccurately* modelled as CCCS. Vbe's what you
really want to look at with BJTs.
 
F

Fred Bartoli

Jan 1, 1970
0
Paul Burridge said:
FETs are certainly accurately modelled as VCCS in a certain regions of
operation. BJTs are *inaccurately* modelled as CCCS. Vbe's what you
really want to look at with BJTs.

Please, tell us more :)
 
D

Dr. Slick

Jan 1, 1970
0
Instead of counting on my memory. I dug out the BLF276's data sheet.

At 100MHz Zin = 1.3 -j1.7 (points taken from graph)

A 1.25Ohms Xl would be very important in this case since it is about as
big as R and Xc. Th package on this is a 6 pins as source configuration.

mag[Z]=sqrt(1.3**2+1.7**2)

so it's still going to be low impedance.

S.
 
D

Dr. Slick

Jan 1, 1970
0
Paul Burridge said:
On 12 Nov 2004 12:30:32 -0800, [email protected] (Dr. Slick) wrote:


FETs are certainly accurately modelled as VCCS in a certain regions of
operation. BJTs are *inaccurately* modelled as CCCS. Vbe's what you
really want to look at with BJTs.


Well, it can be either base current or Vbe, obviously.

But you talk about the "Beta current gain factor" as
(delta Collector current)/(delta base current), and they
don't describe FETs like this.

However, you are correct that "gm" or "transconductance"
can be either dIdrain/dVgate or dIc/dVbe.

The point is that MOSFETs have far less gate current
than do BJTs have base current.


Slick
 
Top