Trio 9R-59DS 'pulling'

[G6Tanuki]

My VFO still has the shroud, yes: I'll experiment with taking it off.

[I know what you mean about convector-heater-caused drift: as a student I lived in a Welsh farmhouse where the entire heating was convectors, and we were at the end of a rather long and ancient DNO feed. You could watch the lights flicker every couple of minutes as someone's convector turned on or off. Not at all good for free-running LO stability !]

[Andrew2]

Originally Posted by G6Tanuki

My mains is quite stable here [I'm the only property on a commercial 3-phase circuit all the way back to the local substation - I can pull 80A and there's only about 0.3V drop!]

What I now see is what I guess you could call 'prolonged warm-up drift' - at 20MHz (receiving the 4th harmonic of a 5MHz xtal) from switch-on there's an initial shift of about 12KHz high over 5 minutes, then it sort-of settles down and continues to drift at about 1KHz a minute after that. Which makes me think it's warm-up related.

I'd rather like *not* to have to chase 20M-band SSB stations across the dial.

That is what mine had. A constant movement in one direction. Well, constant if you ignore the pulling on strong sigs!

[G6Tanuki]

Further experimentation - I found that the leadouts to the second [unused] triode of the ECC85 were unconnected and floating, as was the internal screen [pin 9]. Earthing these seems to have given a slight reduction in warmup drift.

It puzzles me though - why the designers specified an ECC85 - with its unnecessarily powerful 0.45A heater (that's the same power as a 6V6!) - only to use half of it. Something like a 12AT7, with only one half lit - would only take 0.15A and so reduce the amount of excess heat noticeably.

2.83 Watts of heater for the ECC85, 0.945 Watts of heater for half a 12AT7.

(The ECC85 doesn't have a centre-tapped filament so you *have* to power both triodes even if you're only using one.)

My next experiment will be to see which of the half-dozen examples of the ECC85 I have here (Tungsram, Mullard, Valvo) give the least drift.

[Phil G4SPZ]

Maybe they thought that the higher heater power would swamp any ambient temperature changes and improve the drift performance? Is that what made them slip a matt black shroud over it too? Or else the ECC85 was simply a lot cheaper than the 12AT7...

[G6Tanuki]

Unless one of them pops up on-net to give commentary I guess we will never know what went through the minds of the 1960s Trio-Kenwood Corporation designers when selecting their valves. [their use of the 6BE6 - which has a rather high equivalent-noise-resistance - as mixer rather than a 6BA7 or half an ECH81 - is also strange].

But I do have a dozen or so 12AT7WA [the ruggedized, controlled-warmup-heater version specified for various bits of 1950s military gear] here so could do some experimentation.

Now I've got pin-9 grounded it's a dead-simple swap.

[turretslug]

Perhaps they needed the high gm of the ECC85 to guarantee oscillation as L/C ratio got lousy on the highest band, and the good emission from that heating power helped to sledgehammer circuit losses, especially if the tuned elements had been doused in wizzo Japanese damping wax? The Pye CAT here has a modest little EC90 (1/2 ECC82 in effect) with gentle 0.15A heater as 1st LO, things were obviously running fine on the lowest 7 bands but by band 8 (15-31MHz), the get-by measures included shorting most of the anode load to bump anode volts and current up (down to 4k7 from 37k7) and connecting a boost circuit between anode and grid. Obviously, a bit of scratching-around going on.

I'm slightly suspicious of the comment in the manual that using the unused ECC85 section as cathode follower could be "interesting"- it's almost as if they had tried it and run into problems.... I'm scratching my head as to any single triode counterpart of the ECC85 as with the EC90/ECC82, maybe they were too.

Maybe the 6BE6 was "usual" in the big US market and the bean-counters jibbed at the 6BA7? I sometimes have to remind myself when playing around with ex-professional sets that HM/US taxpayer stumped up for the bells and whistles, but with consumer goods, every farthing was being watched like a hawk.

[Radio Wrangler]

Sometimes the choice of devices in one radio is strongly influenced by what parts the firm is using in quantities in its other products. Maybe they already had a part number set up for the ECC85 in their FM receiver product line and had not set up one for any single triode?

You can't reverse-logic device choices by looking at just the one product.

The EC90/6C4 oscillates very nicely in my EA12, and that uses the triode parts of the ECH81 mixers as oscillator buffers

David