[HBR] HBR - BFO progress

[Ron Barlow]

Hi Walt,
 
 Your interesting comments, re parasitic oscillations, reminded me of a long ago incident.
 
Some 40 years ago, I constructed an 80m VF0, using a fairly high transconductance 7 pin miniature tube (6BZ6 or 6CB6, I believe). 
 It had a peculiar problem, that happened to occur in ~ the mid portion of the desired freq range. It would "jump over" a particular frequency (~ 3.6 Mhz, IIRC), that I was attempting to adjust the vfo to!
 The vfo would perform fb, at 3.5 Mhz, and it would continue to function fine, as I increased its frequency, until its frequency approached 3.6 Mhz, then it would "pole vault" over 3.6 Mhz, to ~ 3.61 Mhz. 
 A similar situation occurred when I attempted to adjust the vfo freq, downward, from above 3.6 Mhz. No matter what I tried, the vfo refused to oscillate at 3.6 Mhz!!
 
 The possibility of a rotor contact issue, on the vfo tuning cap was eliminated.
 
 After some had scratching, I began to suspect that a VHF parasitic oscillation was involved. I installed a 33 ohm resistor, in series with the control grid lead, of the oscillator tube. The problem was cured.
 
 It was my guess that 3.6 Mhz was a sub harmonic of the frequency of the undesired VHF parasitic oscillation, that was occurring, simultaneously, with the desired HF oscillation.
 
 After that experience, I began installing parasitic suppressors, as a matter of routine.
 
       73 de Ron  
  

________________________________
From: Walt Hutchens <waltah at earthlink.net>
To: HBR Receiver List <hbr at mailman.qth.net> 
Sent: Sunday, January 5, 2014 5:13 PM
Subject: Re: [HBR] HBR - BFO progress


Ron said:

>  I see a problem with the suggested Colpitts circuit, that you
> provided a link to. If I understand the labeling on the schematic
> diagram correctly, there is a .01 mfd capacitor, on each end of the
> RFC. The cap on the left side of the choke, is electrically in shunt
> with the 180 pf cap, that comprises part of the oscillator tank
> circuit. This will profoundly alter the feedback ratio, and the
> resonant freq of the tank ckt.  In fact, the feedback ratio will be
> altered to the point that I seriously doubt that the circuit will
> oscillate at all.

That's a problematic circuit either way. Without the 0.01 mfd cap it
will oscillate at roughly the frequency set by the tank circuit but
the choke is in shunt with the 180 mmf capacitor. You really don't
want a choke in a VFO circuit: They are terribly unstable.

I've never had any trouble as simple as getting the tap right for a
Hartley. The turns ratio is the same as the proper one for the caps in
a Colpitts circuit. If I remember correctly the command sets use a
separate winding for feedback in that BFO; that's another way to do it
and the turns ratio is the same as the ratio appropriate for a Hartley
or Colpitts circuit.

You can get rid of the choke in a Colpitts circuit by using a resistor
instead and for the very highest Q (hi-Q helps with stability against
tube gain variations due to structural and voltage fluctuations ...)
tap the coil APPROXIMATELY at zero RF voltage, letting the capacitors
set the exact feedback ratio and/or tuning range.

As long as the voltage across the resistor is just a small fraction of
the total the effect on circuit Q will be unimportant.

Feedback ratios really aren't supercritical. The amplitude of
oscillation in a class 'C' oscillator is set by the load imposed when
the grid is driven positive over a fairly wide range of ratios. (The
grid is then basically a shunt diode that eats the excess energy.)
Only if there isn't enough feedback or it becomes so great that the
tube goes into blocking oscillations will there be trouble.

(Blocking: the grid capacitor charges so much during the positive
swing that when the tank circuit swings the grid negative the tube
cannot conduct again even on the next positive swing of the tank.
Instead the capacitor must discharge through the grid leak resistor
for several or many cycles of the RF before the tube again conducts.
This is a good thing for a superregenerative detector, not so good for
a BFO, LO, etc. Cure involves some combination of less feedback, lower
grid leak resistor, and smaller grid coupling cap.)

Well ... long leads and/or the latest UHF-capable tube types can get
you VHF parasitics if you don't take precautions and these parasitics
may occur over just a part of a cycle of the intended oscillation
frequency. But once you've seen this you'll recognize it. Horrific
crashing noise when you tune the oscillator is one symptom because
there's very high VHF current in the capacitor bearings or contacts.

Nearly all construction project articles have clear enough directions
to get you past such issues. It's best to NOT substitute 'better'
tubes than those specified. 6SK7 replaced by a 6BZ6? GOOD chance of
VHF parasitics there, unless you use countermeasures.

This is true throughout the RF-IF stages of a receiver. Also in such
stages 'hotter' tubes routinely need a cross-socket shield to separate
input from output.

The latest tube types will require extra work compared to those of the
50's and if the project you're doing is from a 50's magazine article
the how-to for dealing with the later tubes won't be in the article or
circuit. This can be about the most challenging troubleshooting you'll
see.

All resistors in an oscillator should be the modern carbon film ones:
They're more stable in every respect than the vintage ones. AVOID
bypass-type disk ceramics in an oscillator, even for bypass jobs: They
are NOT stable, indeed their capacitance may vary with the voltage
across them as well as temperature.

Mica bypasses work well in oscillators, even the larger 'postage
stamp' micas.

You're looking for a few parts per million stability, right? Receiver
oscillators are precision instruments: You can breadboard them with
'whatever' parts but use only the best for the actual build.

Walt Hutchens
KJ4KV