lower frequency limits for hf transistors?

[[email protected]]

I've been looking at some specs for uhf transistors (power RF type)
and I notice that not only do they have an upper frequency limit (which
I understand) but there is also a lower frequency limit which I dont'
understand. What is it about these devices that there would be a lower
frequency limit too, eg. they may work well at 450mhz but not
recommmended for 7mhz.

Don WA5NGP

 

[Joerg]

I've been looking at some specs for uhf transistors (power RF type)
and I notice that not only do they have an upper frequency limit (which
I understand) but there is also a lower frequency limit which I dont'
understand. What is it about these devices that there would be a lower
frequency limit too, eg. they may work well at 450mhz but not
recommmended for 7mhz.

Don WA5NGP

Can you post a data sheet where they say that? I have used RF
transistors from DC to wherever I needed. The BJTs just don't have much
beta and you need to maintain RF layout practices. If, for example, you
use a BFS17A as a relay driver and run long lines then it might "sing".

Also, I have used UHF hotrods (AFAIK a BF998) as audio phase shifter
devices. IOW as rather mundane controlled resistors. But don't tell
anyone...

 

[Tom Bruhns]

I've been looking at some specs for uhf transistors (power RF type)
and I notice that not only do they have an upper frequency limit (which
I understand) but there is also a lower frequency limit which I dont'
understand. What is it about these devices that there would be a lower
frequency limit too, eg. they may work well at 450mhz but not
recommmended for 7mhz.

Don WA5NGP

Can you give a specific example? Can you determine if it's a hard
limit, or just a recommendation? The parts I've come across that are
actually limited to some frequency range are more complicated that just
a simple transistor. (I'm thinking, for example, of some integrated
medium power amplifiers from MiniCircuits that are not rated to go
below 10MHz. They really don't, too.)

One problem with using a UHF-rated part at HF is that you will have to
pay very close attention to keeping it from oscillating at UHF. That
is, if it has gain well beyond the frequencies of interest, and you
don't pay attention to what it's doing over the whole range of
frequencies, you can get stung. Clearly, if you are building a
wideband amplifier, you'll need a part you can use over that whole
bandwidth. Some folk build power amplifiers that cover four or more
decades of frequency, up to hundreds of MHz.

Cheers,
Tom

 

[Joerg]

Can you give a specific example? Can you determine if it's a hard
limit, or just a recommendation? The parts I've come across that are
actually limited to some frequency range are more complicated that just
a simple transistor. (I'm thinking, for example, of some integrated
medium power amplifiers from MiniCircuits that are not rated to go
below 10MHz. They really don't, too.)

One problem with using a UHF-rated part at HF is that you will have to
pay very close attention to keeping it from oscillating at UHF. That
is, if it has gain well beyond the frequencies of interest, and you
don't pay attention to what it's doing over the whole range of
frequencies, you can get stung. Clearly, if you are building a
wideband amplifier, you'll need a part you can use over that whole
bandwidth. Some folk build power amplifiers that cover four or more
decades of frequency, up to hundreds of MHz.

One example would be the input amp for lab gear such as network
analyzers or even something as mundane as a frequency counter. They
often need to go from DC to several GHz.

 

[Phil Hobbs]

I've been looking at some specs for uhf transistors (power RF type)
and I notice that not only do they have an upper frequency limit (which
I understand) but there is also a lower frequency limit which I dont'
understand. What is it about these devices that there would be a lower
frequency limit too, eg. they may work well at 450mhz but not
recommmended for 7mhz.

Don WA5NGP

The input and output impedances of RF power transistors tend to be quite
low. Some manufacturers put matching elements inside the package to
help this problem, and of course they work only over a limited frequency
range.

Cheers,

Phil Hobbs

 

[Ancient_Hacker]

I've been looking at some specs for uhf transistors (power RF type)
and I notice that not only do they have an upper frequency limit (which
I understand) but there is also a lower frequency limit which I dont'
understand. What is it about these devices that there would be a lower
frequency limit too, eg. they may work well at 450mhz but not
recommmended for 7mhz.

Don WA5NGP

One possibility:
If you hook up a 450MHz transistor to a 7MHz circuit, the circuit is
likely to have multiple spurious resonances at several places in the
VHF range. Since the transistor has lots of gain at those frequencies,
it's likely to break out into oscillation at the VHF frequencies.

 

[Joerg]

The input and output impedances of RF power transistors tend to be quite
low. Some manufacturers put matching elements inside the package to
help this problem, and of course they work only over a limited frequency
range.

That, and the beta is low. With the exception of MOSFETs, sometimes
called UHF tetrodes for some reason. The BF998 one example:

http://www.nxp.com/acrobat_download/datasheets/BF998_2.pdf

 

[Don Lancaster]

Can you give a specific example? Can you determine if it's a hard
limit, or just a recommendation? The parts I've come across that are
actually limited to some frequency range are more complicated that just
a simple transistor. (I'm thinking, for example, of some integrated
medium power amplifiers from MiniCircuits that are not rated to go
below 10MHz. They really don't, too.)

One problem with using a UHF-rated part at HF is that you will have to
pay very close attention to keeping it from oscillating at UHF. That
is, if it has gain well beyond the frequencies of interest, and you
don't pay attention to what it's doing over the whole range of
frequencies, you can get stung. Clearly, if you are building a
wideband amplifier, you'll need a part you can use over that whole
bandwidth. Some folk build power amplifiers that cover four or more
decades of frequency, up to hundreds of MHz.

Cheers,
Tom

Back in the aerospace daze, I had a broadband 50 MHz amplifier using
2N918's.

Output power meter would get "stuck" at 30 milliwatts with no input.

Using the fanciest test equipment then available, we verified a 950 MHZ
oscillation.

Cured by putting a ferrite bead under one base lead.

--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email: [email protected]

Please visit my GURU's LAIR web site at http://www.tinaja.com

 

[Tom Bruhns]

One example would be the input amp for lab gear such as network
analyzers or even something as mundane as a frequency counter. They
often need to go from DC to several GHz.

Well, yes, although those don't generally have to put out several
watts, which is what I inferred that the OP is looking at.

Cheers,
Tom

 

[Tom Bruhns]

The input and output impedances of RF power transistors tend to be quite
low. Some manufacturers put matching elements inside the package to
help this problem, and of course they work only over a limited frequency
range.

Cheers,

Phil Hobbs

On the other hand, the reactances of anything useful at UHF will
generally be insignificant at 7MHz, and designing the appropriate
matching at 7MHz should not be difficult. Or maybe I should say,
"there are well-known techniques..." Now if the combination of
internal reactances and the external circuit makes the part unstable,
that would be worth considering!

Cheers,
Tom

 

[Joerg]

Well, yes, although those don't generally have to put out several
watts, which is what I inferred that the OP is looking at.

True, and the classic VFETs of the 80's are now history. Was great while
it lasted :-(

Supposedly one of the German ham radio operators (DL9AH?) has published
an article about how to build a huge shortwave power amp using a
gazillion cheap RF transistors.

 

[[email protected]]

OK, I think I've got it. If you use a uhf transistor down at the lower
frequencies you may run into parasitic oscillations at very high
frequencies that could ruin your day (ie burn it up). And you may not
even have test equipment on hand that you would let you observe the
problem if you are way down there on the mhz range.

Makes sense.

tks
Wa5ngp

 

[Joerg]

OK, I think I've got it. If you use a uhf transistor down at the lower
frequencies you may run into parasitic oscillations at very high
frequencies that could ruin your day (ie burn it up). And you may not
even have test equipment on hand that you would let you observe the
problem if you are way down there on the mhz range.

Or have the neighbor banging on your door because the football game on
TV was blurred out by it...

 

[[email protected]]

I guess as they say "I'll just burn that bridge when I get to it".