|
 |
Buffering oscillator for PLL input - RF Cafe Forums
|
Charl
Post subject: Buffering oscillator for PLL input Posted: Sun
Jul 23, 2006 2:21 pm
Colonel
Joined: Fri May 19,
2006 5:01 am
Posts: 25
Location: Netherlands
Hello,
I got my VCO (95-120MHz) working. It is based on the ECC85 tube
and a Hartley configuration (tapped inductor). However, I want to have
a PLL IC (SAA1057) control the frequency. Its feedback input has an
impedance of 135 ohm.
I tried connecting this input to the tap
on the inductor, through a resistor on the resonant circuit directly,
I even tried an additional inductor near the main one to pick up the
oscillations. But the input impedance is just too darn low. I'm afraid
I'm going to have to buffer it, JUST for the PLL IC.
Any suggestions
for a as-simple-as-possible buffer, or some cunning trick to get it
to work without a buffer?
Much obliged,
Charl
Top
nubbage
Post subject: Posted: Mon Jul 24, 2006 4:58 am
General
Joined: Fri Feb 17, 2006 12:07 pm
Posts:
218
Location: London UK
Typically, a low value capacitor, eg
10pF, was used to couple to the following stage. Should the input Z
of the follwing stage be a low value, coupling was taken from the cathode
circuit, where the source impedance was also low.
Passive matching
will always result in no isolation against the effects of load impedance
variation, and the attendant effect on the oscillator stability. Buffers
are used not only to do impedance matching, but also because their high
reverse isolation reduces these adverse effects on the stability.
I have seen high switching speed opto-isolators used as oscillator
buffers (source was Agilent, but I have no record of the type).
Top
Charl
Post subject: Posted: Mon Jul 24, 2006
5:41 am
Colonel
Joined: Fri May 19, 2006 5:01 am
Posts: 25
Location: Netherlands
Dear nubbage,
I tried
connecting the inductor tap (which is connected to the cathode) to the
PLL IC through various caps, 10-1000pF. None of them worked. I think
that with low value caps, the voltage on the PLL input isn't enough
(it needs 10mV). With the high value caps, the loading is too great
and the oscillator cuts out or does weird stuff.
I was hoping
to get away with using a cheap transistor and a few resistors and caps
for the buffer. Any circuit suggestions regarding this?
Thanks
for your time,
Charl
Top
nubbage
Post subject:
Buffering oscillator for PLL inputPosted: Mon Jul 24, 2006 7:52 am
General
Joined: Fri Feb 17, 2006 12:07 pm
Posts:
218
Location: London UK
Hi Charl
One method that used to be
used was to tap the coupling cap a small percentage of the way up from
one end of the resonator coil. Trial and error was the only way of doing
this I recall, but if you have a good RF voltmeter (hi Z input) you
could measure the tap position that gives 10mV. Increase the tap point
away from the common connection to the resonator capacitor until you
have enough voltage, but no further because you start to load the Q
of the resonator until oscillation is wobbly or stops.
For a
buffer I would suggest a dual gate MOSFET 40673 or 3N128 types, that
have reasonably high Zin even at VHF. A source resistor of a few ohms
(non-inductive of course) will increase Zin considerably by negative
feedback.
This should aid with matching to the high Z of the oscillator.
Another tip might be to find an oscillator circuit with a low Z
point.
Maybe a dual triode cascode design?
Top
Charl
Post subject: Posted: Mon Jul 24, 2006 10:52 am
Colonel
Joined: Fri May 19, 2006 5:01 am
Posts: 25
Location: Netherlands
Hello,
Thank you very much for your
comments. I think the best solution to this is to just add a FET as
you described. I don't think I need to be concerned much with clipping/distortion
because it is the input to the (digital) PLL anyway. Am I right about
this?
Also, do you have any design examples or an online reference
where I might learn more about this?
Kind regards and thanks
again,
Charl
Top
nubbage
Post subject: Buffering
oscillator for PLL inputPosted: Mon Jul 24, 2006 11:11 am
General
Joined: Fri Feb 17, 2006 12:07 pm
Posts: 218
Location:
London UK
Hi Charl
I think an IGFET like the RCA 3N128 or
40673 would serve well.
I have a vague feeling there might be a problem
if the gate voltage swing is too high, in that the charge surge into
the gate capacitance at high frequences means the Zin appears highly
reactive, meaning a high capacitance and thus rapidly decreasing impedance.
However, I am sure there are scores of dudes out there with far more
experience in this area than I have.
Re sources: the best would
be amateur radio articles in QST (USA), RSGB (UK) and DARC (Germany)
etc.
Posted 11/12/2012
|
|
|
|
