RF Transistor Selection

[sthim]

Hi,

I am working on a rf preamp project and I've been having trouble
finding a transistor. Some of the transistors are not stable for the
frequencies that I'm working on (2 different preamps for 200MHz and
500MHz) or others do not have the NF I'm looking for (0.5db). I was
hoping if somebody could suggest some transistors... fets, hempts,
anything ...

Thanks
sthim

 

[Frank Raffaeli]

Hi,

I am working on a rf preamp project and I've been having trouble
finding a transistor. Some of the transistors are not stable for the
frequencies that I'm working on (2 different preamps for 200MHz and
500MHz) or others do not have the NF I'm looking for (0.5db). I was
hoping if somebody could suggest some transistors... fets, hempts,
anything ...

Thanks
sthim

if you use a high mobility, high current amplifier, stability may be
difficult near the nearly DC frequencies you mention ;-) . You might
find that, in general, if a transistor is well charactarized at 6 GHz,
its going to be tough as nails to get a stable input match at 200 MHz.
If you tell us the linearity (IIP3), current, input match and gain you
are striving for, it will be easier to make a more specific
recommendation. Try bipolars, JFETs, even large geometry GaAs and SiGe
from Toshiba, NEC(CEL) IBM, Infineon Philips and Agilent. I've had
varying degrees of success with each in different applications. The
toughest challenge will be to try to find a device with NF min at
optimum near the other parameters above. If your input match doesn't
fall within the required noise circle on the smith chart, you may need
to use negative feedback (at the expense of gain and perhaps
stability) to change the input parameters.

Best of luck

Frank Raffaeli
AOM Wireless
http://www.aomwireless.com/

 

[Ricardo Matos Abreu]

I'm not sure sure what IIP3 is.

IIP3 - 3 order Input interception point.

This is a non-linear parameter. For low-power input, for each dB you add at
the input you get +1dB at the output (of course) and +3dB third order
distortion. This translates in a graphic like this:
- x-axis: input power;
- y-axis: output power;
- two lines:
- one for "signal" (1st order), "growing" at 1 dB/dB
- one for "distortion" (3rd order), "growing" at 3 dB/dB

These two lines cross at a given point. The abciss of this point is known as
IIP3, i.e., IIP3 *would* be the input power you would apply in order to get
at the output the same power for both 3rd order distortion and signal.

I wrote *would* because for such a high input power these lines are no
longer straight lines. But for low power they are. One should extrapolate
the lines from low-power in order to achieve the cross point and hence the
IIP3.

Hence IIP3 is related with non-linear 3rd order distortion, which is the
predominant in-band distortion.


Ricardo

 

[skipp hangs from the sonic server]

Cost is going to be an issue here... Generic Gasfets often used in
preamps are your best bet here... but there are some newer devices by
Agilent that are smoking good quality.

Under $25 per device, go with the Gasfets...

Under $160 per device, go with the Aglient Devices. As you can figure,
the IP3 for the Agilent Devices are much better.

But the IP3 for the common ARR used type Gasfet is not that bad. When
you get into "busy places" the better devices pay off. Contact me
through the sonic server Email icons if you need additional information.

200 and 500 Mhz... not a problem... been there, done that, coffee mug and
tee shirt.

cheers
skipp
http://sonic.ucdavis.edu

: Hi,

: I am working on a rf preamp project and I've been having trouble
: finding a transistor. Some of the transistors are not stable for the
: frequencies that I'm working on (2 different preamps for 200MHz and
: 500MHz) or others do not have the NF I'm looking for (0.5db). I was
: hoping if somebody could suggest some transistors... fets, hempts,
: anything ...

: Thanks
: sthim