Recommendation for front-end transistor
Bill Meacham
front-end RF transistor (FET) for a shortwave receiver?
Overload immunity and good IMD are the primary concerns in
this case.
--
Bill Meacham
Vieques, PR
jim bennett
2N4416 for HF
3SK48 HF-VHF
Good performers and cheap.
jim K4KAE
"Bill Meacham" <ex...@coqui.net> wrote in message
news:39D1012A...@coqui.net...
Steve Ratzlaff
U310/J310 (plastic case) have the best overload/IMD characteristics of
any jfet available, but you may have trouble just "sticking" it into
your receiver rf amp circuit without rebiassing the jfet. Idss of these
are in the 25-35mA range; the device likes a nominal 10-12mA actual bias
to achieve those overload/IMD characteristics.
Steve AA7U
Glen Ross
Bill Meacham <ex...@coqui.net> wrote:
> What would be the group's recommendation for a good
> front-end RF transistor (FET) for a shortwave receiver?
> Overload immunity and good IMD are the primary concerns in
> this case.
>
The RF stage is not going to suffer from these problems [assuming the
various element voltages are near correct for the device] because the input
signal levels are so low. IMD and similar effects occur much later in the
chain where the signal levels can easily run into several volts.
Glen G8MWR
Martin Ackroyd
Glen Ross wrote:
>
[snip]
> >
>
> The RF stage is not going to suffer from these problems [assuming the
> various element voltages are near correct for the device] because the input
> signal levels are so low. IMD and similar effects occur much later in the
> chain where the signal levels can easily run into several volts.
I think this is debatable. Intermodulation in the front end by
strong unwanted signals at nearby frequencies can be the major factor
limiting performance. Because most of the selectivity of a receiver
comes from the later stages, these have not yet been eliminated.
Some sets [eg Racal RA17] incorporate an attenuator so that the
overall input signal can be reduced in amplitude in cases where
cross modulation is a problem.
Cheers
Martin
Glen Ross
Martin Ackroyd <martin....@virgin.net> wrote:
> Some sets [eg Racal RA17] incorporate an attenuator so that the
> overall input signal can be reduced in amplitude in cases where
> cross modulation is a problem.
>
You miss the point<G> The attenuator is placed at the input but the
whole point of it is to reduce intermod caused by high signal levels in
later stages. The 1st RF stage in an RA17 can handle several volts of
input with no IMD/cross-mod problems.
I hate RA17s; having had the job of refurbishing over 300 of the things for
the MOD.
Glen G8MWR
Eamon Skelton
>front-end RF transistor (FET) for a shortwave receiver?
>Overload immunity and good IMD are the primary concerns in
>this case.
>Bill Meacham
>Vieques, PR
It all depends what what your requirements are. A
dual gate mosfet will give you high gain, low noise
and reasonable IMD performance.
The BF961 is cheap to buy and easy to find.
Others: BF981, 3SK88,3SK45, 40673.
A grounded gate J-FET amp will have much less
gain, low noise and will probably have better
IMD performance than than the dual gate FET.
The 2SK125 is a popular choice.
A bipolar UHF power transistor is still hard to
beat. The two popular configurations are:
1
Common emitter with plenty of negative feedback.
See Solid State Design for the Radio Amateur.
This arrangement works very well in the front
end of the Elecraft K2 transceiver. You can
download the schematic from their web site.
2
Common Base with negative feedback to the
emitter via a single turn winding on the
output transformer (Norton Amplifier).
Use a 2N5109 or 2N3866 with a heat-sink
and lots of current.
Thats my opinion for what it's worth. I would go
for the bipolar transistor unless you want the
ultimate in strong signal handling which is
probably a power mosfet. G3SBI published designs
for an ultra linear receiver using a VMP4 mosfet
as the RF amp and four mosfets as a switching
mixer (H-Mode mixer).
Remember that there is little or no advantage
to be gained by building an RF amp that has
a higher output intercept point than the
input intercept of the following mixer.
73, Ed. EI9GQ.
--
Remove 'X' to reply.
EI9GQ homebrew radio page
http://www.qsl.net/ei9gq
Lawrence E. Stoskopf
of the transistors are as advertised in the posts, however, I'm still a
bit confused on the basics:
Agree that there is no purpose in using a "stronger" transistor than the
following mixer,..... Use of a VMP-4 running hundreds of milliamps
certainly assures that it will not get out of its "linear" range
quickly. But how linear is that linear range? That device has multiple
wires to the die and seems to "act" better than some equivalent single
bonded devices. But we are discussing strong signal performance.
Then Motorola and others have "Ultralinear" transistors with multiple
bonds that seem to be advertised for CATV applications where multiple
signals of equivalent strength need to be amplified. Are these more
linear than the FET devices?
And again, some of the MRF150 series devices are advertised as having
low higher order distortion, but really run a lot of current.
So for my front end, I would like a very linear device at reasonable
overload properties (shouldn't have a real big signal there) running a
reasonable current. Haven't really seen that comparison in the stuff I
read. Most is simply what is available easily.
N0UU
Bill Meacham
My mission was to beef up both the front end and mixer of a
garden variety late 60s/early 70s SWL receiver. The main
problem at this QTH is a 5kw AM station about 1/4 mile away.
So while some may struggle for lower noise, others may want
more gain, and others better IMD/Overload performance.
Achieving all three perfectly most likely results in some
degree of compromise, but in the case of a late 60's rig, I
suspect better performance can be achieved in all 3 areas
for about a buck or two.
Eddie Haskell
used them for EVERYTHING,from hi-Z amps in their multimeters to front ends
in the "HW" series receivers.I remember replacing them often.
Jim Weir
shared these priceless pearls of wisdom:
->The 40673 if I remember correctly was quite fragile to overload. I remember
replacing them often.
I've got several tens of thousands of them in the field in the front end of a
VHF transceiver(s) operating in high-RF field mobile environments. Other than a
couple of infant mortalities (probably as much to do with rug-shuffling on a
cold day than anything) once in the circuit and soldered down, they are damn
near bulletproof.
I thought RCA was the only vendor on this device. I've got a few hundred of
them left in overstock, and was saving them for the ultimate end of life
replacement on the transceivers. Is there a currently producing manufacturing
source for them (i.e. I'll use them in new designs unless the only stock is the
dusty pile of them left from the RCA days)?
Jim
Jim Weir, VP Eng. RST Eng. WX6RST
A&P, CFI, and other good alphabet soup
Ian Jackson
<j...@rst-engr.com> writes
Anybody tried a BFY90 as a bipolar frontend? It was a good 'small-
signal' work-horse for wideband CATV amplifier 1st stages.
Note that beefier 2N3866, although often used in early wideband CATV
transistor amplifiers, was NOT designed for class-A operation. It is a
multiple-emitter device WITHOUT ballast resistors, and is essentially a
class-C device, relying on skin effect to share the emitter currents
equally. Class-A operation, at fairly high currents, caused certain
problems as the transistor aged. Later CATV transistors (2N5109?) were
specifically designed for ultra-linear class-A.
Ian.
Martin Ackroyd
I'd be glad if you'd help me out. Maybe I have some basic misunderstanding.
Glen Ross wrote:
>
> Martin Ackroyd <martin....@virgin.net> wrote:
>
> > Some sets [eg Racal RA17] incorporate an attenuator so that the
> > overall input signal can be reduced in amplitude in cases where
> > cross modulation is a problem.
> >
>
> You miss the point<G> The attenuator is placed at the input but the
> whole point of it is to reduce intermod caused by high signal levels in
> later stages.
Doesn't the use of agc mean that the signal at the later stages is
controlled to an appropriate amplitude (and, also, by the time it reaches
later stages, interfering signals away from the desired frequency will
also have been attenuated by the slectivity of the tuned circuits and/or IFT's?
[eg with a big enough agc voltage, the valves will actually attenuate,
rather than amplify]
Isn't it quite possible to have volts of interfering signal at the
grid of the 1st rf stage? [eg with transmitters operating relatively
nearby]?
Hope you can help my understanding improve.
Cheers,
Martin
Michael Black
> Hi Glen,
>
> I'd be glad if you'd help me out. Maybe I have some basic misunderstanding.
>
> Glen Ross wrote:
> >
> > Martin Ackroyd <martin....@virgin.net> wrote:
> >
> > > Some sets [eg Racal RA17] incorporate an attenuator so that the
> > > overall input signal can be reduced in amplitude in cases where
> > > cross modulation is a problem.
> > >
> >
> > You miss the point<G> The attenuator is placed at the input but the
> > whole point of it is to reduce intermod caused by high signal levels in
> > later stages.
>
> Doesn't the use of agc mean that the signal at the later stages is
> controlled to an appropriate amplitude (and, also, by the time it reaches
> later stages, interfering signals away from the desired frequency will
> also have been attenuated by the slectivity of the tuned circuits and/or
IFT's?
> [eg with a big enough agc voltage, the valves will actually attenuate,
> rather than amplify]
>
> Isn't it quite possible to have volts of interfering signal at the
> grid of the 1st rf stage? [eg with transmitters operating relatively
> nearby]?
>
> Hope you can help my understanding improve.
>
> Cheers,
> Martin
>
>
The problem isn't that amplifier stages won't overload, but that mixers
will overload first. An amplifier only has to amplify.
But a mixer is mixing, and is by it's function more prone to overload
than a straight amplifier.
I'm not going to use units, just numbers. Lets say there is a 1 level
at the input to the first amplifier, and it provides a gain of 10. Assuming
it doesn't overload, it's output will be 10. So the mixer that follows
is seeing ten times what the amplifier saw at its input. And the mixer
is easier to overload, so it suffers from that and the fact that the signal
is amplified in the amplifier. Take out the amplifier, and the mixer
can handle stronger signals at the antenna, but of course has to deal
with the lack of amplification that might be helpful on weaker signals.
Concern about the amplifier overloading is only a problem if the mixer
can handle the signal out of the amplifier. If that mixer can handle
that 10 level, then you can start wondering if the amplifier will
overload on strong signals.
Another way to look at it is that if the amplifier after the antenna is
overloading, you can be reasonably sure that the mixer is overloading
also, at least in simple receivers.
If you put an attenuator between the antenna and the amplifier stage,
yes, you are reducing the signal level that the amplifier sees at it's
input. But if you reduce the signal that the amplifier sees by 1/2,
the output of the amplifier will be only at level 5 for the same level
1 at the antenna, and the mixer may be better able to handle that.
I don't know where AGC fits into this, but certainly you do want selectivity
as close to the front-end as possible. That's why you should put a good
filter right after the mixer, so the following stages do not have
to handle more than the bandwidth that the signal occupies. You do
see amplifiers before crystal filters, in receivers that are using
diode-ring mixers, but the amplifiers are intended to deal with strong
signals without overloading, by use of proper devices and proper biasing.
You need that amplifier there in that case, because of the losses
in the mixer. But you want as little amplification as possible before
whatever is used to provide ultimate selectivity, something that wasn't
always the case in receivers of days gone by. And whereas at one
time you would have significant gain before the mixer, some of
those old tube receivers had two stages of RF amplification, you now
want as little amplification before the mixer as possible.
The ideal receiver would put all the selectivity between the antenna
and the first stage. But there are problems with that, which of
course is why we have superheterodyne receivers.
Michael VE2BVW
Glen Ross
blac...@cam.org (Michael Black) wrote:
> The problem isn't that amplifier stages won't overload, but that mixers
> will overload first. An amplifier only has to amplify.
> But a mixer is mixing, and is by it's function more prone to overload
> than a straight amplifier.
>
Thanks, Michael. What was I was trying to convey was that the RF stage was
the least likely stage to suffer from cross-mod. One of the most "bomb
proof" Rx I ever used was an old Atlas job with no rf stage at all. Signal
went through a BPF and straight into a diode ring mixer. That thing could
handle anything you threw at it.
I agree that some RX designs have used power fets and such things in the
1st stage. The only things of this type that I have been involved in the
design of were to be used co-sited with high power TX systems. That is a
completely different ball game from what is found in most domestic use.
Glen G8MWR
Lawrence E. Stoskopf
you stay in the "Linear" operating point of a device, how do you know
how linear the "Linear" range us? Surely there must be some distortion
with any device, even though it is no measurable or is meaningless.
Splitting hairs.
N0UU