High gain output valve

[Neil Purling]

I want a high-gain valve I can use to give a speaker output from a detector stage. It does not have to produce a high power output.
It ought to have 6.3V heaters. Whether it is B7G IO or B9a, it does not matter.

[TonyDuell]

Does it have to be one stage, or could you use something like a triode pentode which would give you 2 stages in one 'bottle'? One of the ECL8x series, for example.

[paulsherwin]

The EL84 is the obvious one to go for - still in production, plentiful and relatively cheap if you stay away from fashionable brands. Most output valves will produce reasonable output volume from the volt or so at the detector though, so you might as well use whatever you have to hand.

[G8HQP Dave]

6CL6? 6CH6/EL821? Video output valves generally have higher gm than similar size audio output valves.

[Neil Purling]

Got plenty of EL84, Reflektor & Svetlana. I'll leave the Telefunken & Lorenz examples to those with far too much money.
I have used 6CH6, I thought that apart from some pin differences they were about the same as a EL84. The human ear is such a poor guide.

[G6Tanuki]

Historically there are plenty of high-gm valves that were used in "short" superhets: the EL33 is a pentode with a gm of 9 and was sometimes driven direct from the detector.

The EBL21 was also popular in cheap superhets at one time [double-diode-pentode so you could do detector/AGC and audio output in one bottle], it has a brother in the EBL31.

The Pen45DD could be worth a look too: http://www.r-type.org/exhib/aad0289.htm

Video output-valves often have spectacular gm [the EL821 has already been mentioned]. As audio stages these valves can be problematic - at heart they aspire to a second life as oscillators and will do all they can to 'take off' in inconvenient ways.

If you want real fun, look at the battery-powered 1D8GT: a diode-triode-output-pentode all in one bottle! http://www.radiomuseum.org/tubes/tube_1d8gt.html

[Neil Purling]

I may have an example of the 6P25, maybe a EL33.
As far as sensitivity goes they would be arranged thus: EL84, EL33/6P25 and then the 6CH6.
Not a lot in it, if gm is the whole story in judging the potential gain to be had from any valve.
If we were considering a simple bedside radio, it does not have to fill the whole room against any competing noise.

[paulsherwin]

You will find just about anything will be adequate. If you just want half a watt or so (plenty for a bedroom radio) you can even use a non traditional valve like an EF80 or EF91.

[Herald1360]

For something off the wall, try an EF80 or EF91. Reasonable gm and enough anode diss capability (2.5W) for about a watt of output.

Edit: I think there's an Oxford effect here somewhere

[Neil Purling]

I have seen an example of an application that used all EF91's back when they were ten a penny as pulls from old TV's.
The regenerating detector varies the screen boltage untill it is just short of howling.
Surely something there which is very high gain would be difficult to control.

[Herald1360]

Varying the screen voltage effectively changes the gain to whatever is needed just not to hoot. It can be much tamer to do that than mess around with the degree of feedback via variable capacitors or inductor coupling.

[Skywave]

The use of an ECL86 might be worth a try: triode pre-amp. and O/P pentode in one valve.

The Racal RA-17 uses an EF91 for audio O/P to a 3-Ohm spkr. Quite adequate volume for a small / medium-sized room.

Al.

[Silicon]

The Racal output transformer has a ratio of 93:1 to match the 3 Ohm speaker to the 25K Ohm anode load impedance of the EF91.

You could use a standard 40:1 output transformer with a 15 Ohm speaker.

[Peter.N.]

I have seen an EF80 used as sound output in a commercial set, there must be thousands of them floating about.

Peter

[Radio Wrangler]

More Gm with EF184 and still 2.5W anode rating.

David

[bikerhifinut]

ECL86 are hard to find now and pricy.
Maybe ECL83? They are usually to be found quite cheap and plenty good enough.
I have a few new unused PCL86, but they need a 14V 300mA heater (They will work ok at 12.6v). You'd be welcome to one to try I didnt pay a lot for them.

A.

[cmjones01]

I can remember my first attempts to add a loudspeaker amplifier to a regenerative TRF detector years ago. The first valve I used was an AC2Pen/DD which did exactly what it said on the tin. No instability and adequate volume. I think the Gm is reasonably high because it's a valve intended for short superhets. However, the 4V heater and large physical size of the pre-war AC2Pen/DD didn't fit with the rest of the project, so I changed to an EL84 which had noticeably more gain but was much more difficult to work with. My straggly wiring made the whole thing very prone to oscillate. Of course, if I'd built it properly, I wouldn't have had that problem.

Chris

[Neil Purling]

As for the ECL86: I believe it is because the design was a very late one it is not easy to find 'old stock' examples. How many valve manufacturers produced the ECL86 I don't know. The Russians don't seem to have made any, don't know about the Czechs or in E. Germany.

It is a good idea to use an EF91 as o/p. The RS Universal o/p transformer has a maximum Primary load of 15K ohms. Nowhere near being a good match.

Now. Is the gm figure for a valve the sole indicator of gain? I want to know in order to select the valve for the regenerating detector and the output It would suggest that a pair of EF91 as the detector and the o/p would be a good choice,. But would that be the correct choice?

[Skywave]

Originally Posted by Neil Purling

Is the gm figure for a valve the sole indicator of gain?

No: the 'gain' of a valve - in terms of amplification factor, µ - can be expressed as the product of the mutual conductance, gm, and the a.c. anode resistance of the valve, ra, or put symbolically, µ = gm x ra.
If the valve 'sees' a load of RL, then the amplification of the stage, A = (µ. RL) ÷ (ra + RL).
Hence the 'working gain' of the valve is reduced from its amplification factor, µ, when a load is connected. And that is simply because the load, RL, is effectively in parallel with ra

For a clear explanation of all that, refer to 'Foundations of Wireless' by M.G. Scroggie, 6th. edition, p.127.

Al.

[Skywave]

Originally Posted by Neil Purling

I want to know in order to select the valve for the regenerating detector and the output It would suggest that a pair of EF91 as the detector and the o/p would be a good choice,. But would that be the correct choice?

Generally speaking, I have been led to believe that for a 'good' regenerative detector, a valve with a low gm is required. On that basis, a valve like the 6BR7 would seem to be a good choice, unlike the EF91. As for the output stage, gain and power output capability, the above post of mine refers.

Al.