6J5 or one half of 6SN7 as an output stage?

[Neil Purling]

Is it viable to use a 6J5 or half a 6SN7 as an output device as I don't need much power?
Just wondering about the values of the g1 resistor and the size of the cathode bias resistors.
If more gain is needed I have a 6SL7.

[G4XWDJim]

The R107 military receiver uses an EBC33 as the output valve very successfully with good volume.

A 6J5 will do very well.

Jim

[Radio Wrangler]

You haven't said what sort of output. Are you wanting an RF output for transmitter? An audio amp to drive headphones? An audio amp to drive a loudspeaker?

David

[paulsherwin]

You can use most valves for class A output if you only need a couple of hundred mW. There were lots of hobbyist designs using cheap TV valves like the EF80 or EF91.

[Neil Purling]

To drive a small speaker, (4" x 2.5" elliptical 8 ohm) probably via a RS Universal o/p transformer. The only other small speaker I have came from a portable that had a DL9X as audio o/p. Not sure what power we are going to get from the 6J5 or the drive level it will need.

[G6Tanuki]

A 6J5 has a rated plate-dissipation of 2.5W and maximum cathode-current of 20mA.

With 250V HT, biasing to 10mA cathode-current will mean it's dissipating 2.5W.

Of that, assuming class-A, you will get a theoretical maximum of 1.25W of audio out; when you introduce real-world losses you'd be more likely to get 0.75W of usable audio.

[Neil Purling]

So: More than enough for a DIY radio.
What of the grid and the cathode resistor's values?
I have looked at the circuit of the R107, but could not see any component values.

[G6Tanuki]

Grid-resistor - well, since it's not going to be passing any current (well, it shouldn't be...) and you're not depending on it for 'contact bias' then the value is pretty flexible - I'd suggest anything between 100K and 470K Ohms.

Cathode-resistor, well, if you go here: http://www.r-type.org/pdfs/6j5-1.pdf

you can download the spec sheet which includes curves showing the current for different grid-bias voltages and anode-voltages. Work out the grid-bias voltage you need to get, say, 10mA of current flowing. Then a simple bit of Ohm's Law will tell you the cathode-resistor value you need to generate that voltage when 10mA is passing through it!

[Looking at the curves, to get 10mA anode current at 250V you need to be running about 8V of grid-bias, meaning a cathode resistor of 800 Ohms]

Remember to bypass the cathode resistor with a nice big electrolytic; given how compact these are nowadays I'd go for something around 47uF.

[Herald1360]

A low impedance triode like ECC82 or something with similar characteristics but less audiophile appeal will work best. You want a valve that needs enough negative bias when set to dissipate near its Pa max that you can roughly double the anode current without going into grid current. A load impedance around 8k should get you a few hundred mW of output power. A bit more than a typical '50s valve portable......

Of course with a double triode there's always class B push-pull to play with though you might need to be up for rewinding an output transformer.

[short wave]

R107 circuit at..
http://bama.edebris.com/manuals/britmil/r107
regards S-W

[ms660]

RCA tube book gives some useful info on a triode configured for a power output valve, starts on page 28:

http://www.tubebooks.org/tubedata/RC25.pdf

Lawrence.

[Leon Crampin]

If you use a 6SN7 you can parallel the two triodes for a greater output to good effect. The output transformer ratio will have to be carefully calculated as most transformers for pentode use will have a higher ratio than is optimum for a low impedance triode.

Leon.

[barrymagrec]

If you are using a double triode such as a 6SN7 or an ECC82 and don`t need the other half you can parallel the two halves and get more power or reduce the loading of each valve.

EMI did this for the line output of the BTR2 tape machine.

[kalee20]

The 6SN7 was used commercially as an output valve in the Soundmirror tape recorder (both halves used in push-pull). It works well.

I have used a 6J5 as a single-ended output valve. Did you know it is virtually pin-compatible with the 6V6 (obviously not including the screen-grid!)? There's not masses of volume, but it does work effectively.

Choice of operating point is a bit tricky with a triode. Best plan is to use a high anode voltage (250-300V rather than 200V), then lots of grid bias to throttle the Ia down to what you need (bearing in mind the 2.5W max anode dissipation). Then you can apply more input signal without running into grid current.

[Radio Wrangler]

The 6J7 is not far removed from half an ECC82 or half a 6SN7.

The power you can get depends on the HT available, and also the impedance of the speaker and the ratios of available output transformers, otherwise you might not be able to approach what the valve data says is possible.

David

[turretslug]

Originally Posted by kalee20

I have used a 6J5 as a single-ended output valve. Did you know it is virtually pin-compatible with the 6V6 (obviously not including the screen-grid!)? There's not masses of volume, but it does work effectively.

The manual for the CR100 actually suggests substituting an L63/6J5 for the KT63/6F6 output valve to reduce HT loading when running from batteries, taking advantage of this common pinning across many IO output valves.

[Sideband]

A few years ago I was playing about with a simple superhet. I called it 'minimalist' because it didn't have an IF amp and was really cut to the bone. In that I was restricted with the mains transformer size that I had and eventually used an ECC82 as output valve and AGC rectifier....the second triode strapped as a diode. However using one half as output valve, a cathode resistor of 820 ohms, an HT of around 220V and grid leak at 470k, gave me an anode current of around 8.2mA (ECC82 rated at 10mA per triode). Plenty of output into a small 5" round speaker and using an RS universal output transformer set to highest ratio (it didn't seem to make a lot of difference) it was too loud to run at full volume in an average room.

[TonyDuell]

My Agaphone wire recorder uses half an ECC33 to drive a speaker. It doesn't give room-filling sound but it is certainly audible and would be OK for an experimental radio or similar.

Didn't at least one well-known TV (HMV??) use an EF80 as the audio output valve?

[Radio Wrangler]

Racal RA17 used an EF91 to drive its small speaker. It was OK but not good. THe RA117 upped the audio output valve to an EL91

David

[Jon_G4MDC]

And that could drive the RA117 front panel LS up to capability + a bit.