"Dinky" AF Output Stage

[Bazz4CQJ]

I want to build a valve output stage which is frugal in terms of HT and LT current, space and weight. I think that 0.5 to 1 watt input would do and quality is not too crucial. HT is around 180-200V and LT is 6.3V. Input level would be from an existing high impedance headphone feed. Searching the net has turned up quite a lot of quite recent projects where people have built comparable amps to use for guitar practice! These seem to be often based on the likes of ECC81's, either single ended, single ended but in parallel and in push-pull. Older ideas include ECL82's and EF91's among others. Most output transformers intended for valves seem to be intended for use at appreciably higher power levels but I wonder whether a small ferrite torroidal (lightweight) component could be found / used?

Any insightful suggestions on good ways to do this would be welcome.

Bazz

[Anthony Thomas]

I am only a bungling amateur but I do know that many output transformers have an "air gap" in their laminations, this is I believe to avoid the core becoming saturated by the dc content of the HT supply. Choose an appropriate output transformer would be my advice. I am not sure if a torroid would work due to the air gap or lack of, some other knowledgable person could say if this is so. There is also the question about core material to take into consideration.

A good choice of valve would be the ECL82 I would think, the least suitable valve would be perhaps the EF91 because it is not realy designed as an output valve. The ECL82 is a better choice if you have a choice.

I have seen a suggestion to use a small mains transformer as an output transformer by restacking the laminations and placing a piece of paper between the end laminations to give an "air gap" effect. I have not tried it but have modified one transformer to one day experiment with this and use a 6V6GT output valve.

Tony

[RF Burn]

That old workhorse the ECL80 springs to mind as being capable of fitting your requirements, they are abundant and cheap, require 170 - 200V DC supply at about 22mA, and provide about 1 watt output.

Just make sure that the common cathode pin on the valveholder is connected as directly as possible to the ground/chassis/0V using a good quality capacitor of at least 470uF (16V+), otherwise this valve tends to become a very good oscillator. If splitting the cathode resistor in order to supply the correct bias for the triode grid then you may need to decouple that point too with a second capacitor.

I know that there is a very good reason for NOT using ferrite as the core material for audio power transformers, it may be due to saturation but I can't remember exactly what it is. However suitable output transformers appear regularly on the auction sites at reasonable prices, for the ECL80 you require a 10K - 12K primary impedance, most of the small old Radiospares (RS) multi-ratio transformers are suitable.

Adrian

[G8HQP Dave]

You could try an ECC82 in push-pull. No airgap needed in the OPT, but you would need a phase splitter. You could use another twin triode for this, using any of the common circuits.

[paulsherwin]

To avoid the stability problems you could just use the pentode section of an ECL80 and ignore the triode. There should still be enough gain to do what you want, and the circuit would be very simple.

[G6Tanuki]

Given your low HT, I'd suggest looking at some of the valves used in US radios intended to run off 110V AC/DC supplies - these are designed to pass quite high currents with low anode-voltages and could easily get you a Watt or so if run "single-ended". 25L6 ?

If you want to go push-pull, a 12BH7 would be worth looking at.

[turretslug]

Also the commonplace EL91/6AM5/CV136, low heater current, good for a watt or so single ended and also used in push-pull in a few Eddystone sets- which will give ideas for phase-splitters and interfacing generally.

[paulsherwin]

There are a number of output valves designed for car radio use which have frugal heaters - I can think of the EL42 and the EL85, and I'm sure there are many others. Unfortunately they aren't very common.

[kalee20]

I agree that this is crying out for an ECL80. Low power requirements, adequate sensitivity, just watch the common cathode.

Slightly better quality could be expected from an ECC82, both halves in push-pull, but as pointed out you will need another, acting as phase splitter.

As for a ferrite toroidal output transformer, I have never seen one or heard of this being used. In principle, you could use a ferrite toroid as a push-pull output transformer, but why? Better to use a toroidal iron one, with 5 times more flux density before saturation, thus allowing a much smaller transformer for the same spec. Ferrite would come into its own at higher frequencies, such as if you wanted to drive an ultrasonic transducer, but forget it for audio.

As a single-ended transformer, forget it - the DC standing current will saturate the core and you'll have very disappointing performance. Unless, you use an anode choke and parallel feed the transformer via a capacitor. But then, you are back into the argument in the above paragraph.

[Synchrodyne]

ECC40?

Also, there is a 2 x ECL80 circuit in Mullard "Valves, Tubes and Circuits" #2 & 3.

Cheers?

[Bazz4CQJ]

Thanks for the various suggestions. Searching through my stash, I do seem to have acquired more ECL80's than I could ever account for, and a couple of reasonably suitable OPT's that could be used with them, so I think I'll start off down that route. There's a lone ECL82 that needs testing in reserve.

I find increasingly that when I search through the treasure that I find things I have no recollection of ever seeing before; I am never sure whether this is a consequence of the increasing stock acquired over the last 50 years, or the decreasing number of operative neurons now available .

[julie_m]

If it's any help, I've been building a little portable amplifier using an ECL80 powered from a 6V rechargeable battery, using a solid-state vibrator for the HT (+120V). I tied the + side of the battery to 0V and used a couple of preset pots to give me the necessary negative bias on the grids, sidestepping the shared cathode. I'm using an RS multi-tapped output transformer; primary pins 1 and 3 seem to give the most volume. It draws 1A from the battery, of which 0.3A is powering the heater.