Looking for a transistor suggestion, for an RF receive pre-amplifier to
cover 10Hz to about 60Mhz. So far, I have found a pair of 2N5109B bipolar
transistors, which work ok in a grounded base push-pull design. These
transistors provide a gain of about 12db over an input voltage range of 1
microvolt or less to over 4 volts. These transistors were designed many
years ago; hopefully better transistors now exist which can provide superior
linearity and better isolation form input to output. In addition, a spice
model must be available.



Thanks.

On Nov 2, 10:01 pm, "numeric" wrote:
Looking for a transistor suggestion, for an RF receive pre-amplifier to
cover 10Hz to about 60Mhz. So far, I have found a pair of 2N5109B bipolar
transistors, which work ok in a grounded base push-pull design. These
transistors provide a gain of about 12db over an input voltage range of 1
microvolt or less to over 4 volts. These transistors were designed many
years ago; hopefully better transistors now exist which can provide superior
linearity and better isolation form input to output. In addition, a spice
model must be available.

Thanks.

Some of the best transistors for that sort of application are those
designed for CATV amplifier service, where low distortion and low
noise are important. Careful and creative circuit design can help a
lot, too.

Your 4V input suggests you get 16V out at a gain of 12dB. Is that all
at 50 ohms, and RMS? If so, that's about five watts output.

Have you measured IIP3 for your circuit? If so, what sort of
performance do you get? What sort of noise figure? Can you share
more details about the circuit?

Do you really mean input-to-output isolation, or output-to-input
isolation? Normally, I'd consider +12dB in the forward direction to
be poor "isolation" but in a direction I didn't really want isolation
anyway--I want gain that way. You should be able in a push-pull
circuit to add parts in a cross-coupling manner to lower the output-to-
input isolation considerably, if that's what you are after. You can
also run each side cascode (or in your case more literally, grounded-
base into grounded-base...), and that can help the linearity and noise
(as the input stage will run at lower voltage drop and thus run
cooler), as well as the isolation.

I asked here not all that long ago about ideas for a low-gain, low-
noise preamp with high IIP3 (and of course even higher OIP3) -- my
mention of +55dBm as an IIP3 goal met with some disbelief (and one
very respected suggestion for a way it could be done with some parts
no longer readily available), but my thought is that it should be
possible, given a fairly high power dissipation. Some of the modern
op amps are getting close to being able to do it, at least through HF
(30MHz); not to 5 watts output though. Not sure why you need that in
a receiver. ;-)

Cheers,
Tom

"K7ITM" wrote in message
ups.com...
On Nov 2, 10:01 pm, "numeric" wrote:
Looking for a transistor suggestion, for an RF receive pre-amplifier to
cover 10Hz to about 60Mhz. So far, I have found a pair of 2N5109B
bipolar
transistors, which work ok in a grounded base push-pull design. These
transistors provide a gain of about 12db over an input voltage range of
1
microvolt or less to over 4 volts. These transistors were designed many
years ago; hopefully better transistors now exist which can provide
superior
linearity and better isolation form input to output. In addition, a
spice
model must be available.

Thanks.

Some of the best transistors for that sort of application are those
designed for CATV amplifier service, where low distortion and low
noise are important. Careful and creative circuit design can help a
lot, too.

Your 4V input suggests you get 16V out at a gain of 12dB. Is that all
at 50 ohms, and RMS? If so, that's about five watts output.

Have you measured IIP3 for your circuit? If so, what sort of
performance do you get? What sort of noise figure? Can you share
more details about the circuit?

Do you really mean input-to-output isolation, or output-to-input
isolation? Normally, I'd consider +12dB in the forward direction to
be poor "isolation" but in a direction I didn't really want isolation
anyway--I want gain that way. You should be able in a push-pull
circuit to add parts in a cross-coupling manner to lower the output-to-
input isolation considerably, if that's what you are after. You can
also run each side cascode (or in your case more literally, grounded-
base into grounded-base...), and that can help the linearity and noise
(as the input stage will run at lower voltage drop and thus run
cooler), as well as the isolation.

I asked here not all that long ago about ideas for a low-gain, low-
noise preamp with high IIP3 (and of course even higher OIP3) -- my
mention of +55dBm as an IIP3 goal met with some disbelief (and one
very respected suggestion for a way it could be done with some parts
no longer readily available), but my thought is that it should be
possible, given a fairly high power dissipation. Some of the modern
op amps are getting close to being able to do it, at least through HF
(30MHz); not to 5 watts output though. Not sure why you need that in
a receiver. ;-)

Thanks for the CATV transistor suggestion. Getting data for RF transistors
should be easy, but after searching the Internet, the ARRL handbook, QST and
other sources it's proven elusive. Part of the problem is companies that
made transistors, such as Motorola, no longer make these types of devices;
they sold the part of the company that made all the good stuff.



Quickly answering your questions:

The design goal for the RF pre-amp can provide several watts output,
although if the RF path is preceded with band-pass filters, the likely hood
of actually finding signals with volt magnitude levels is small.

The voltages are peak-to-peak; it's easy to set spice source parameter
voltages using peak-to peak values. The isolation is from output to input.
Any voltage change appearing at the output, i.e. due to mixer products,
should not appear at the input (antenna port). The gain is small, yet large
enough to overcome mixer and filter losses.

As previously indicated, the design is push-pull. The base is grounded for
each half. The emitters are biased through a center-tapped transformer
with -4volts through a 50-ohm current limiting resistor. The collector is
similarly powered through a center tapped multi-winding transformer with +12
volts (actually need higher voltage to support the extreme limits). The
Internet http://www.qrp.pops.net/preamp.asp has a single ended example for a
"A Low Noise, High Dynamic Range Broadband RF Amp".

The design is still in the simulation stage using spice software. Linear
Technology LTspice SwitcherCadIII has been particularly useful, but also
used other spice simulation software. I started by simulating a design
published in QST (Feb, 1993) to explore the differences between spice
simulation and a proven design. The spice simulation did affirm the proven
design. Any differences found, were traced to details the author left out.
In addition, the spice models had to be written or find an existing model;
such as the 2N5109B spice model. The transformers were simply modeled as
inductors with mutual coupling. LTSpice supports a very useful mutual
coupling syntax, where Kxxx L1 L2 [L3 ...] coefficient value for values
between -1 to +1. For realistic transformers using a value of less then one,
can yield low and high frequency roll-off characteristics that match real
world RF transformers.

"numeric" wrote in message
...

"K7ITM" wrote in message
ups.com...
On Nov 2, 10:01 pm, "numeric" wrote:
Looking for a transistor suggestion, for an RF receive pre-amplifier to
cover 10Hz to about 60Mhz. So far, I have found a pair of 2N5109B
bipolar
transistors, which work ok in a grounded base push-pull design. These
transistors provide a gain of about 12db over an input voltage range of
1
microvolt or less to over 4 volts. These transistors were designed
many
years ago; hopefully better transistors now exist which can provide
superior
linearity and better isolation form input to output. In addition, a
spice
model must be available.

Thanks.

Some of the best transistors for that sort of application are those
designed for CATV amplifier service, where low distortion and low
noise are important. Careful and creative circuit design can help a
lot, too.

Your 4V input suggests you get 16V out at a gain of 12dB. Is that all
at 50 ohms, and RMS? If so, that's about five watts output.

Have you measured IIP3 for your circuit? If so, what sort of
performance do you get? What sort of noise figure? Can you share
more details about the circuit?

Do you really mean input-to-output isolation, or output-to-input
isolation? Normally, I'd consider +12dB in the forward direction to
be poor "isolation" but in a direction I didn't really want isolation
anyway--I want gain that way. You should be able in a push-pull
circuit to add parts in a cross-coupling manner to lower the output-to-
input isolation considerably, if that's what you are after. You can
also run each side cascode (or in your case more literally, grounded-
base into grounded-base...), and that can help the linearity and noise
(as the input stage will run at lower voltage drop and thus run
cooler), as well as the isolation.

I asked here not all that long ago about ideas for a low-gain, low-
noise preamp with high IIP3 (and of course even higher OIP3) -- my
mention of +55dBm as an IIP3 goal met with some disbelief (and one
very respected suggestion for a way it could be done with some parts
no longer readily available), but my thought is that it should be
possible, given a fairly high power dissipation. Some of the modern
op amps are getting close to being able to do it, at least through HF
(30MHz); not to 5 watts output though. Not sure why you need that in
a receiver. ;-)

Thanks for the CATV transistor suggestion. Getting data for RF transistors
should be easy, but after searching the Internet, the ARRL handbook, QST
and
other sources it's proven elusive. Part of the problem is companies that
made transistors, such as Motorola, no longer make these types of devices;
they sold the part of the company that made all the good stuff.



Quickly answering your questions:

The design goal for the RF pre-amp can provide several watts output,
although if the RF path is preceded with band-pass filters, the likely
hood
of actually finding signals with volt magnitude levels is small.

The voltages are peak-to-peak; it's easy to set spice source parameter
voltages using peak-to peak values. The isolation is from output to input.
Any voltage change appearing at the output, i.e. due to mixer products,
should not appear at the input (antenna port). The gain is small, yet
large
enough to overcome mixer and filter losses.

As previously indicated, the design is push-pull. The base is grounded for
each half. The emitters are biased through a center-tapped transformer
with -4volts through a 50-ohm current limiting resistor. The collector is
similarly powered through a center tapped multi-winding transformer with
+12
volts (actually need higher voltage to support the extreme limits). The
Internet http://www.qrp.pops.net/preamp.asp has a single ended example for
a
"A Low Noise, High Dynamic Range Broadband RF Amp".

The design is still in the simulation stage using spice software. Linear
Technology LTspice SwitcherCadIII has been particularly useful, but also
used other spice simulation software. I started by simulating a design
published in QST (Feb, 1993) to explore the differences between spice
simulation and a proven design. The spice simulation did affirm the proven
design. Any differences found, were traced to details the author left out.
In addition, the spice models had to be written or find an existing model;
such as the 2N5109B spice model. The transformers were simply modeled as
inductors with mutual coupling. LTSpice supports a very useful mutual
coupling syntax, where Kxxx L1 L2 [L3 ...] coefficient value for values
between -1 to +1. For realistic transformers using a value of less then
one,
can yield low and high frequency roll-off characteristics that match real
world RF transformers.



Just a thought - even if not blending in with the scheme of the Homebrewing
aspect.

There are a couple pre-amps made to sell which boast a KHz to several
hundred MHz of frequency response in amplification ability. Maybe you could
check them out. Hosfelt Electronics - I believe - carries them.

Also - I have some older Motorola Data Books here - if you need any info on
a particular item. If nothing else, maybe get something to come close to
matching.

Maybe not much help but - thought I'd toss it in.

Lou

On Nov 2, 9:01 pm, "numeric" wrote:
Looking for a transistor suggestion, for an RF receive pre-amplifier to
cover 10Hz to about 60Mhz. So far, I have found a pair of 2N5109B bipolar
transistors, which work ok in a grounded base push-pull design. These
transistors provide a gain of about 12db over an input voltage range of 1
microvolt or less to over 4 volts. These transistors were designed many
years ago; hopefully better transistors now exist which can provide superior
linearity and better isolation form input to output. In addition, a spice
model must be available.

Thanks.

He did ask for a RECEIVE pre-amp, to 60mHz
Not a power amp, and not 60GHz, and you don't need 4volts for a
receive preamp

How about and MPF102 JFET in Class A
Bias it at about 8mA, and you should get ~20db power gain

On Nov 5, 7:43 am, Joe wrote:
On Nov 2, 9:01 pm, "numeric" wrote:

Looking for a transistor suggestion, for an RF receive pre-amplifier to
cover 10Hz to about 60Mhz. So far, I have found a pair of 2N5109B bipolar
transistors, which work ok in a grounded base push-pull design. These
transistors provide a gain of about 12db over an input voltage range of 1
microvolt or less to over 4 volts. These transistors were designed many
years ago; hopefully better transistors now exist which can provide superior
linearity and better isolation form input to output. In addition, a spice
model must be available.

Thanks.

He did ask for a RECEIVE pre-amp, to 60mHz
Not a power amp, and not 60GHz, and you don't need 4volts for a
receive preamp

How about and MPF102 JFET in Class A
Bias it at about 8mA, and you should get ~20db power gain

Because there are those of us who, for a "receive preamp", would very
much like to get IIP3 greater than +50dBm, and ideally up around
+60dBm. It may only need to handle 0dBm or +10dBm signals, but it
does need to have very low distortion. Not going to happen with an
MPF102, at least not without an exceptionally creative circuit around
it. I'm not quite sure why the OP wants a preamp in the first place,
since noise figure on HF is generally not that much of an issue. It
may be in part to get isolation between the mixer and the antenna so
he doesn't radiate the LO signal, but that can be done with tuned
circuits at least for narrowband radios. I have slightly different
reasons for wanting a modest gain very low distortion amplifier like
that. In any event, it would be a nice circuit to have in my "bag of
tricks."

Cheers,
Tom

"K7ITM" wrote in message
ups.com...

How about and MPF102 JFET in Class A
Bias it at about 8mA, and you should get ~20db power gain

Because there are those of us who, for a "receive preamp", would very
much like to get IIP3 greater than +50dBm, and ideally up around
+60dBm. It may only need to handle 0dBm or +10dBm signals, but it
does need to have very low distortion. Not going to happen with an
MPF102, at least not without an exceptionally creative circuit around
it. I'm not quite sure why the OP wants a preamp in the first place,
since noise figure on HF is generally not that much of an issue. It
may be in part to get isolation between the mixer and the antenna so
he doesn't radiate the LO signal, but that can be done with tuned
circuits at least for narrowband radios. I have slightly different
reasons for wanting a modest gain very low distortion amplifier like
that. In any event, it would be a nice circuit to have in my "bag of
tricks."

You are correct, the pre-amp is needed to get isolation between the mixer
and antenna. I want a very high dynamic range and broadband design, as wide
as possible to cover the LF to 60Mhz range. I have simulated several mixers,
including H-mode mixers; they all disturb the antenna input unless a good
preamp is used. The low noise figure is needed mainly for the VHF range and
yet the performance comes free in the LF to HF range.

Although I haven't completed reviewing all suggestions, I do appreciate the
comments from all.

Thank You.