6V6s on their sides, and negative-feedback.

[TonyDuell]

I am pretty sure there was an Ultra TV with the line output valve (30P4?) and booster diode (U191?) mounted horizontally. The model VP17/72 springs to mind, but I may be totally mis-remembering that.

Going on a little, sitting behind me is a B&O colour television that I must get round to restroing. The line output valves (PL504 and PL509), boost diodes (PY800A and PY500) and vertical output valve (PL508) all sit horizontally. I am pretty sure the sound output valve (can't remember the number, but it's B9A based) does too.

[MotorBikeLes]

All the valves in the Grundig hybrid CTVs were side mounted. Models 717, 1500 and 3010. From memory they had PCL85 vertical, PCF802 (line osc) and of course the 2 in lop stage (PY500A, PL509). Can't remember if sound was valve or solid state, but there was a PL95 in the mono set of the era.
Les.

[Bobsound]

The RCA receiving tube manual says that the 6V6/6V6GT can be mounted in any position.

Bob

[Herald1360]

If you have a 6BW6 design you're happy with, just sub 6V6 directly.

QQV03/10 for two channel audio is problematic- the common cathode will need to be very well bypassed or grounded with separate bias supplies. For single channel p-p the cathode is less of an issue but the common g2 rules out UL operation. Using two with both halves in parallel would be easier.

Properly done, NFB will tend to flatten and widen the frequency response. For stability reasons it would be better to do any response tweaking outside the loop at the front end.

[G6Tanuki]

The 6BW6 design I've got was intended as a single-ended AM modulator for an aircraft transmitter - 315V on the anode, 280V on the screen - that's a bit more than the 6V6 is intended for, but Brimar said it was OK for the 6BW6 - in 1950s military aircraft you expected to be shot down before the valves in your transmitter exceeded the "Intermittent Mobile Service" [IMS] service lifetime.

I'm now planning on running my 6SN7/6V6 amp with no NFB, and seeing how it goes. NFB means less gain...

If I experimented with a single 3/10 for stereo, "Ultra-linear" is irrelevant; surely that's only for push-pull stuff??

OK on the issue of bypassing the common cathode but a modern electrolytic will have at most an Ohm or so of ESR and if I was to put a modern 0.1uF Polyester/Polystyrene cap in parallel it'd ensure the cathode was essentially grounded even up to 20/30KHz? Ferrite-beads suitably deployed on the control-grids, and 50- or 100-Ohm wirewounds in each anode would prevent the 3/10 achieving its innate ambition to be a VHF power-oscillator.
I'll be using copious numbers of ferrite beads and 10KOhm series-grid-feed resistors in my 6SN7/6V6 amp in any case - this is a RF-enriched household and a couple of hundred Watts of SSB on 14MHz gets into all sorts of inconvenient places.

Just remember - I'm not talking "Hifi" amplifiers here; I'm talking "Loud". "Does it go up to 11?" is more important than better-than-0.1% distortion.

[G6Tanuki]

Originally Posted by Bobsound

The RCA receiving tube manual says that the 6V6/6V6GT can be mounted in any position.

Thanks!! This is what I'm depending on!

[joebog1]

No feedback, radiogram transformers.
A wonderful recipe for a grungy guitar amplifier,
and DEFINITELY not Hi-Fi.

Joe

[Hartley118]

Well with a pentode and no feedback, you'll get that high output impedance which tends to emphasise the treble as the speaker impedance rises at high frequencies. However, if you were to connect the 6BW6 as a triode, you could claim an audiophool amp with nice sounding even harmonic distortion and a decently low output impedance. Suddenly you're into 'high order' audio! Need more power? Connect two triodes in parallel.

Martin

[G6Tanuki]

I've never really got the whole "connecting-tetrodes/pentodes-as-triodes" thing [except when doing something like a pair of zero-bias triode-connected 807s to get 120-Watts of AF with low-component-count. Apparently doing the same but with two 813s can get you 650 Watts].

Surely the whole point of developing beam/kinkless-tetrodes/pentodes was to overcome the intrinsic limitations of triodes? I am no audiophile who "listens to the decibels", if anything I watch the anode-current meter

[Hartley118]

Well I don't claim to be any sort of expert in valve output stages, but basically a separate screen grid fed from its own HT supply magically renders the valve's anode current relatively independent of anode voltage. In other words, a tetrode or pentode has a much higher output impedance than a triode, as well as typically a higher available output voltage swing and hence higher power capability. When feeding a typical speaker, with an impedance rising with frequency, the high output impedance of a pentode or tetrode results in treble emphasis, hence the common use of a 'tone correction' capacitor. Negative feedback lowers the output impedance, so eliminates the need for tone correction,whilst also reducing distortion.

When it comes to distortion, the tetrode or pentode characteristic tends to flatten the waveform at top and bottom, which generates odd harmonics, generally thought to be unpleasant. Distortion in a triode, however, which doesn't bottom as sharply as a pentode, is more progressive and tends to produce asymmetry in the output waveform, typically progressively compressing the negative half more than the positive top half. This asymmetry generates even harmonics which are not only relatively unobjectionable, but at high frequencies can give a subjectively brighter sound quality.

Of course, push pull operation is well known for its ability to cancel even harmonics , restoring symmetry to the waveform, so I can understand the enthusiasm for a single ended triode output which leaves those even harmonics uncancelled
Martin

[Bazz4CQJ]

Here we are; 15 and 20kHz 6V6 amps from Brimar Data Book#4.

B

[G6Tanuki]

From my time spent designing broadband RF linear-amps with outputs up to a few Kilowatts - the one thing I always made a point of was providing a stable, well-decoupled screen-supply! he idea that it should wobble in any way associated with the signal was anathema if you wanted to minimise IMD.

The "simple" way was to feed the screens from a supply regulated by glow-tubes (strings of VR105 or VR130). For some valves [the EIMAC 4X150/350 kind) it was important to design screen-supplies that could both source *and sink* current.

G3SEK/GM3SEK has a range of good designs for this: http://www.ifwtech.co.uk/g3sek/board.../tetrode-1.htm

"Ultra-linear" operation of tetrode audio amps is really wrong. . . .

[greg_simons]

Originally Posted by HamishBoxer

I am sure in my 45years of servicing I came across LOP valves horizontal but which chassis?? Pye , Thorn come to mind.

yes indeed, the pye 691. And 697. Series used the lop valves horizontally, not a bad old chassis either.

[PJL]

The Pye V14C has a vertical chassis with all valves mounted horizontally.

[joebog1]

At least 3/4's of Australian B/W TV chassis were "fold down from vertical" and ALL bottles were horizontal. ER so was the picture tube

Joe

[joebog1]

Baz4cqk ! this thread started off about single ended designs, rather than push/pull. Its very easy to achieve those sorts of top end frequency responses with PP, rather more difficult with single ended designs, even with a fancy transformer!.
BUT its spiked my interest, and as I have plenty of transformers and valves in this class I have been designing an amp exactly like this for myself.
Watch this space!!

Joe

[Bazz4CQJ]

Joebog! Yes, I realised after I took the time and trouble to scan and upload the circuits that Tanuki has limited himself to SE, but I am pleased to see my efforts were not completely wasted if I have succeeded in inspiring you, and the circuits may be of possible interest to others?

Something that I have been wondering about in the context of '20kHz amps' is the issue that the average age of people using is forum is probably very much on the wrong side of 50, so however good the source or the amp, I wonder just how many (few?) of us can hear anything remotely approaching that frequency ?

B

[joebog1]

Thats OK Bazz,
After all, when you reread the Brimar data book, it reminded you of all sorts of things that you were "meaning to do".

Since this thread started, I have been going through ALL my magazine articles to find suitable data for Tanuki, and trust me I have gigs and gigs worth!!!

Anyway, I remember reading this many years ago, because its specifically designed around second hand parts, or more specifically radiogram parts that are not so fi. Its so similar in every way to a multitude of others thats its really not that important, other than to satisfy that niggle I had remembered about "radiogram transformers".
So what follows only took two weeks of delightful dreaming back and 2 seconds to copy-- Yes nurse, I did take my medication

As I mentioned previously, the power supply is not critical!. It can be derived from a voltage doubler, a bridge rectifier, or a full wave sand rectifier, and as Tanuki wants minimal space used, thats the way I would go. A valve rectifier can also be used, with its BIG increase in heat output from the amp, and another heater supply, unless a modern minature type like a 6V4, although I would use a 6CA4. Anyway, the article is a good read, and although it doesnt go deeply into negative feedback, its does discuss it sufficiently to get the amp going. Should Tanuki require assistance setting up negative feedback, Im sure he will have heaps of help from those here.

Joe

[G6Tanuki]

Thanks for the offer of circuits and specs - I too have plenty! From 1940s ARRL and RSGB handbooks through the various RCA/Mullard/Philips designs and also a collection of Practical Wireless Circuits.

When my hearing was last tested (2012) I had response to a smidgen beyond 18KHz, but tailing off rapidly below 300Hz - the only benefit I can think of from being loudly exploded-at in my mid-20s is that ever since I've been nonresponsive to 50Hz hum!

[G6Tanuki]

My basic amplifier will be something like this:

http://www.r-type.org/articles/art-164.htm

but with half a 6SN7 for each channel as the preamp, and without the feedback/tone-control (tone-control's provided by the signal-sources)

I'll be using some slight component changes too: the grid-resistor of the 6V6 in the circuit shown is 1MOhm; that seems way high to me (220 or 330K is what I tradifionally associate with a 6V6) - there's also a noticeable absence of grid-stopper resistors in both the preamp and 6V6 stages - I always wire a 10K resistor in series with the grid in these sorts of situations so - in conjunction with the grid-to-cathode capacitance - a low-pass-filter effect is formed which reduces the risk of the thing 'taking off' at RF or inadvertently demodulating/reproducing RF picked up on the signal-wiring.

A 10KOhm/100uF R-C network feeding the HT to the anode of the first stage will be included too; the design shown doesn't appear to make any attempt to decouple the HT to the first stage at all!

Perhaps I'm paranoid, but I consider every Amplifier as being an Oscillator that hasn't yet worked out how to achieve its life-ambition!