J M Noeding wrote:
On Sat, 22 Nov 2003 18:20:46 -0500, Kenneth
%wa2mze%@%bellsouth%.%net% wrote:



Not all fets function well into the vhf range. If you want to build a
solid state dipper using fets make sure you get vhf units, not ones
intended for audio use. There are circuits using fets that simulate
negative resistance (like tunnel diodes) which make for an interesting
oscillator since no coil taps are needed and a single gang cap can be used.


Right. I don't like fets, execpt as switches, RX RF amplifier and in
PA. We had a discussion in a radio club among designers 30 years ago
if it was any reason for using fets in oscillators instead of
bipolars, but couldn't really find any good reason, except they were
fancy.
It is also a circuit using a jfet and pnp-combination which works in a
negative impedance mode, but usually found on lower frequencies. Used
it for some time to check resonnant frequency of unknown tuned
circuits with a counter, but the problem is to control oscillation
level. It worked fine for 50% of the applications while the counter
indicated some higher (or lower) frequency for some circuits, so it
was abandomed. Believe I have seen it mentioned for griddipmeters,
too.
It seems to be a reason why PNP UHF transistors are used in most UHF
tuners found here, usually BF479, BF679, BF979. I've copied some of
the text for DL7QY's dip meter covering up to 1400MHz on
http://home.online.no/~la8ak/m3.htm

-JM


--
remove ,xnd to reply (Spam precaution!)
The negative resistance fet circuit I mentioned used a pair of jfets,
one P channel and one N channel with the sources tied together and the
gates of each transistor tied to the drain of the other. In this way
each transistor acts as the self bias resistor of the other. The
resulting voltage/current curve of the combo exhibits a region where the
current decreases as the voltage increases, looking much like the 'kink'
in the plate circuit curves of an old time tetrode tube (such as the
24A). The major disadvantage of this circuit for a gdo is that it is
such a strong oscillator that you don't get much of a dip (you need to
sample the output and feed it to a rectifier and meter) unless you load
the coil down with a shunt resistor....and you need a different value
for each range. Also P channel fets with good vhf performance are as
rare as hen's teeth.

On Sat, 22 Nov 2003 22:02:03 -0500, Kenneth
%wa2mze%@%bellsouth%.%net% wrote:

remove ,xnd to reply (Spam precaution!)
The negative resistance fet circuit I mentioned used a pair of jfets,
one P channel and one N channel with the sources tied together and the
gates of each transistor tied to the drain of the other. In this way
each transistor acts as the self bias resistor of the other. The
resulting voltage/current curve of the combo exhibits a region where the
current decreases as the voltage increases, looking much like the 'kink'
in the plate circuit curves of an old time tetrode tube (such as the
24A). The major disadvantage of this circuit for a gdo is that it is
such a strong oscillator that you don't get much of a dip (you need to
sample the output and feed it to a rectifier and meter) unless you load
the coil down with a shunt resistor....and you need a different value
for each range. Also P channel fets with good vhf performance are as
rare as hen's teeth.

Interesting. The circuit I rebuilt my old Tradiper with used two
dual-gate MOSFETs., 3SK88s I believe they are. It gives a nice fragile
ouput which is easily pulled down by an external tuned circuit,
thereby giving rise to a good, deep dip. The only problem is the size
of the blasted snifffer coils! Whilst this is fine for big old valve
type constructions, it *is* totally impractical for today's ever
smaller PCB jobs. I guess the time is coming to abandon it in favour
of some other method of measurement, probably involving a scope and
sig-gen.

--

"I expect history will be kind to me, since I intend to write it."
- Winston Churchill

On Sat, 22 Nov 2003 22:02:03 -0500, Kenneth
%wa2mze%@%bellsouth%.%net% wrote:

remove ,xnd to reply (Spam precaution!)
The negative resistance fet circuit I mentioned used a pair of jfets,
one P channel and one N channel with the sources tied together and the
gates of each transistor tied to the drain of the other. In this way
each transistor acts as the self bias resistor of the other. The
resulting voltage/current curve of the combo exhibits a region where the
current decreases as the voltage increases, looking much like the 'kink'
in the plate circuit curves of an old time tetrode tube (such as the
24A). The major disadvantage of this circuit for a gdo is that it is
such a strong oscillator that you don't get much of a dip (you need to
sample the output and feed it to a rectifier and meter) unless you load
the coil down with a shunt resistor....and you need a different value
for each range. Also P channel fets with good vhf performance are as
rare as hen's teeth.

Interesting. The circuit I rebuilt my old Tradiper with used two
dual-gate MOSFETs., 3SK88s I believe they are. It gives a nice fragile
ouput which is easily pulled down by an external tuned circuit,
thereby giving rise to a good, deep dip. The only problem is the size
of the blasted snifffer coils! Whilst this is fine for big old valve
type constructions, it *is* totally impractical for today's ever
smaller PCB jobs. I guess the time is coming to abandon it in favour
of some other method of measurement, probably involving a scope and
sig-gen.

--

"I expect history will be kind to me, since I intend to write it."
- Winston Churchill