TV preamp FET choice

[Paul Burridge]

Hi all,

I've decided to build a TV preamp just for the hell of it. There's one
channel on 850Mhz I can't quite pick up without one, anyway. I've got
three JFETs in my junk box that *might* make the grade at this
frequency, but I'm not sure. The parts I have are the BF244, the J310
and J211. Is there going to be any worthwhile gain from any of them,
even in common-gate mode, at the required frequency, or should I be
better off looking at GaAsFETs?

p.

 

[John Larkin]

Hi all,

I've decided to build a TV preamp just for the hell of it. There's one
channel on 850Mhz I can't quite pick up without one, anyway. I've got
three JFETs in my junk box that *might* make the grade at this
frequency, but I'm not sure. The parts I have are the BF244, the J310
and J211. Is there going to be any worthwhile gain from any of them,
even in common-gate mode, at the required frequency, or should I be
better off looking at GaAsFETs?

p.

The easy way would be to use a GaAs or SiGe MMIC. They are wideband
(many GHz) and have nf's around 3 dB. For better noise performance, a
tuned PHEMT can get below 1 dB, for a lot more work. Sirenza has some
very low noise MMICS, and NEC has nice fets. I'd suspect a jfet would
be pretty wimpy up here.

John

 

[Ken Smith]

Hi all,

I've decided to build a TV preamp just for the hell of it. There's one
channel on 850Mhz I can't quite pick up without one, anyway. I've got
three JFETs in my junk box that *might* make the grade at this
frequency, but I'm not sure. The parts I have are the BF244, the J310
and J211. Is there going to be any worthwhile gain from any of them,

The lead inductance of the J310 and J211, I think, rule them out right at
the start. IIRC both these devices have largish gate capacitances.

What you really need for this application is a parametric amplifier and a
stirling cycle device on the idle load.

Anything short of that won't give the performance you want.

 

[Active8]

The lead inductance of the J310 and J211, I think, rule them out right at
the start. IIRC both these devices have largish gate capacitances.

What you really need for this application is a parametric amplifier and a
stirling cycle device on the idle load.

Anything short of that won't give the performance you want.

!!!

 

[Fred Bartoli]

The easy way would be to use a GaAs or SiGe MMIC. They are wideband
(many GHz) and have nf's around 3 dB. For better noise performance, a
tuned PHEMT can get below 1 dB, for a lot more work. Sirenza has some
very low noise MMICS, and NEC has nice fets. I'd suspect a jfet would
be pretty wimpy up here.

Didn't he said that his prototypes invariably end like this
http://cjoint.com/?iAr2yjfIRa ?

 

[John Larkin]

Didn't he said that his prototypes invariably end like this
http://cjoint.com/?iAr2yjfIRa ?

Do you think that would be OK at 800 MHz?

John

 

[Fred Bartoli]

Do you think that would be OK at 800 MHz?

I'm sure he's able to try.

 

[Paul Burridge]

Didn't he said that his prototypes invariably end like this
http://cjoint.com/?iAr2yjfIRa ?

Not quite, Fred. I aspire to make a *final* circuit as neat as that
one.

 

[Paul Burridge]

The lead inductance of the J310 and J211, I think, rule them out right at
the start.

I assume you're referring to the T092's internal bonding wire
inductance?

IIRC both these devices have largish gate capacitances.

Just 2 and 4 pF for the J301? That seems manageable, surely?

What you really need for this application is a parametric amplifier and a
stirling cycle device on the idle load.

"sterling cycle device on the idle load"?? WTF does that mean?

 

[John Woodgate]

I read in sci.electronics.design that Active8 <[email protected]>

!!!

That would need to be connected to the antenna with hoaxial cable, as
well.

 

[John Larkin]

I read in sci.electronics.design that Active8 <[email protected]>


That would need to be connected to the antenna with hoaxial cable, as
well.

Or just pole-mounted.

John

 

[Henry Kiefer]

Use a low-cost BF998 or similar in SMD case, can be 1dB NF up to 1GHz. I can
help you if needed.

cu -
Henry

 

[John S. Dyson]

Hi all,

I've decided to build a TV preamp just for the hell of it. There's one
channel on 850Mhz I can't quite pick up without one, anyway. I've got
three JFETs in my junk box that *might* make the grade at this
frequency, but I'm not sure. The parts I have are the BF244, the J310
and J211. Is there going to be any worthwhile gain from any of them,
even in common-gate mode, at the required frequency, or should I be
better off looking at GaAsFETs?

The J310 type FETS need too much matching to be wideband at
and provide much gain at low noise.

I STRONGLY suggest using the HP 54143 type PHEMT. It biases similarly
to a BJT (needs a small, 0.5-0.7V positive voltage on the gate), and
can provide wideband low noise at any frequency range that it makes
sense to use.

The 54143 can provide a low noise figure with 50-75 source impedance,
and with some matching it can do damned good. It isn't so 'hot' as
to oscillate too easily (but will still probably be one of the hottest
components that you have ever used.) Just make sure that the source
leads (two of them) are very short -- actually, you need a slight
amount of inductance in the source. You'll probably also need some
parallel feedback (1.5K and 150nh in series between the drain and
gate, of course with DC blocking), and that will help alot with matching.
For the output circuit, you might want to spoil the output a little with
some parallel resistance and some series inductance/resistance. This
can help to make the gain stage incredibly stable.

The J310 would probably give you a very small gain at 850MHz withcircuit with <1dB NF. You'll need to bias it at 2-3V@60ma for optimal
performance. The 54143 will likely be nearly overload proof when
compared with the J310. (The 54143 at 60ma can put out 20-50mw, with
<<1dB NF!!!)

On the other hand, some Sirenza components can give you <2dB NF with
wideband gain (e.g. 4586.)

John

 

[John Larkin]

On the other hand, some Sirenza components can give you <2dB NF with
wideband gain (e.g. 4586.)

John

Their 3586 is one of the few "50 ohm" mmics that actually has a 50 ohm
input impedance; most are a lot lower. In fact, you can tweak the
supply current to hit exactly 50 ohms.

John

 

[Paul Burridge]

Use a low-cost BF998 or similar in SMD case, can be 1dB NF up to 1GHz. I can
help you if needed.

Thanks, Henry. It's nice to know someone is prepared to come all the
way from Berlin to climb onto my roof. Gimme your phone number.

 

[Paul Burridge]

Their 3586 is one of the few "50 ohm" mmics that actually has a 50 ohm
input impedance; most are a lot lower. In fact, you can tweak the
supply current to hit exactly 50 ohms.

TV systems are 70 ohms feed in l'l Britland, John. Plus, what has the
supply current to do with anything??

 

[John Larkin]

TV systems are 70 ohms feed in l'l Britland, John. Plus, what has the
supply current to do with anything??

75 ohms here, usually. Do you use 280 ohm twin lead?

Tweaking the supply current of Darlington-type MMICS (the most common
kind) changes the input impedance. Just thought that was interesting.
Most RF guys don't much care; they consider 1.5:1 to be a pretty good
VSWR. I work with stuff where a 2% reflection is a problem.

John

 

[Active8]

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This is also know as "fire for effect", and it seems to work

Or just pole-mounted.

Is it better to mount a pole or mount to a pole?

 

[Paul Burridge]

Use a low-cost BF998 or similar in SMD case, can be 1dB NF up to 1GHz. I can
help you if needed.

Actually, I've just turned up some 3SK88 DG-MOSFETs. Would they
suffice?

 

[Paul Burridge]

75 ohms here, usually. Do you use 280 ohm twin lead?

Erm, no. 70 ohm coax...

Tweaking the supply current of Darlington-type MMICS (the most common
kind) changes the input impedance. Just thought that was interesting.
Most RF guys don't much care; they consider 1.5:1 to be a pretty good
VSWR. I work with stuff where a 2% reflection is a problem.

Indeed. I dare say the gain of the amp if designed properly would
dwarf any mismatch loss between 70 ohms and 50, anyway.