mystery frequency instability in an AVO all-wave oscillator

[Al (astral highway)]

I've just mended and restored an AVO all-wave oscillator (1959 vintage) and have checked its performance, which is generally astounding.

Most frequencies across the range up to 40 Mhz are stable to three and sometimes four places after the decimal point.

However, at the lowest range, frequencies are relatively unstable and rapidly drift by up to a k/c above or below the set frequency. Any theories to explain this very selective anomaly?

thanks

[6AL5W-Martin]

Question:
The low frequency range is also made by Coils, or ist this a seperated part what works only with Caps and Resistors?

If yes, the Pot is a part of the R-C Network and may be the problem ( to test by swap with a metal resistor) it can be also an old cap, which needs to be replaced by a newer type.

[Al (astral highway)]

yes, there are six ranges and six coils in a copper-screened turret.

[Skywave]

Hi Astral.

I'm not too familiar with that particular S/G, but it would seem that the problem is restricted to that band only. Since there is a unique coil assy. associated with that range - which probably has freq.- determining capacitor/s associated with it - that seems to be logical place to start looking. You may have a faulty coil - loose winding, corroded connection, something mechanical - adrift. Could be a leaky cap. - and possibly not one that is part of the freq.- determining components. May be a R that has gone very high and changing the operating point of the osc. valve.

What is the Fmax and Fmin of that particular range? A few kc/s drift in a range of, say, 10 - 50 kc/s is a different issue to the same drift in 10 - 30 Mc/s range, for example.
This data will help with a remote diagnosis like this.

Al / G8DLH

[hilitevr]

Hi Astral

Almost certain to be a fault with the relevent coil, these are wound with cotton covered litz wire which has all the requisites for absorbing moisture thus lowering the Q and eventually causing corrosion. As you might know, litz wire consists of multi strands each of which are enamelled and insulated from each other. when corroded some of these multi strands can become intermitently O/C. In the old days providing corrosion hadn't started we used to bake coils in an oven for a while, mostly worked wonders.

[Al (astral highway)]

Yes, that's the turret there with the copper screen removed. I think I have slightly more recent interconnections but otherwise it's identical.

Thanks for the litz wire theory. It does make sense. Although the inside of the tester was in generally superb condition, moisture was able to collect incrementally in confined places like the dummy aerials, causing destructive and 100% corrosion in the termination of the RFC, for example, and green oxidisation of the brass-copper junctions there. One grub-screw was so corroded that it would not release after a week of being soaked in WD40, and I have had to make a new dummy aerial.

Elsewhere, moisture would have been able to disperse more readily, but again,
condensation could have formed inside the turret and gradually permeated the
litz wire of the lowest-frequency coil (the others are all superbly stable, as I said in my OP, to three or four places after the decimal point!)

So thanks - good theory and I'll maybe apply a gentle hair-drying to the coil in several sessions to see how I get on. If that fails, I may remove and lightly grill the turret ( ) - not!

[Al (astral highway)]

Hi Al, and once again thanks for your helpful suggestions - not easy without a circuit diagram and your RC theory in the absence of that was a good shot (in fact the tuned circuit has a single, dual-ganged condenser of the type used in broadcast receivers and individual coils each with a feedback coupling.

The Fmin of the range is 95K/c and Fmax is 300K/c. In this case, given the selectivity of the fault, I think hilitevr has a winning theory with his litz-wire moisture permeation suggestion.

All the other ranges are exceptionally stable, given the vintage of the tester. In the megacycles, it'll remain stable within 1-5kc on all ranges.

Thanks again for sharing your thoughts, Al.

[Skywave]

Not that it matters, but the "RC theory" wasn't mine, but 6AL5W's!

I was "homing in" on the coil / Litzendraht wire idea, (re: my earlier Post), but hilitevr has beaten me to it! Again, not that it matters! I feel sure that that's where the trouble is - but it is probably worth spending a little time checking out the other (less likely but more "fixable") possibilities first.

If it's any consolation, I (and probably many others here) have had similar problems like this with high inductance Litz wire coils. A "favourite" is the dreaded "green spot" on the surface of the windings of an IFT

I have never found an effective fix to this problem (apart from component replacement) ~ so if anyone has any ideas to help Astral (and myself and others) with problems like this with Litz wire coils - let's be hearing from you - preferably in a new Thread, e.g. "fixing Litz wire coils".

Al / G8DLH

[Al (astral highway)]

I've taken the turret out and am putting it into a higher ambient temperature for a few days. After that, I'll test the faulty LC combination for instability again. If it's better after a few days, that suggests that any moisture hasn't permeated many layers of the coil, and I may as well then seal the cotton with lacquer on the spot, preventing a recurrence. Clearly, I don't want to do this is moisture has slowly permeated deep into the coil's layers.

If it's still a problem, I'll try putting it into a slightly hotter environment for a few days to weeks and then seal it.