Two Valve + IC Regenerative RX kit

[Tony1951]

Not sure if construction projects are allowed here, but here goes...


I just built a funny little valve / IC hybrid regenerative kit I bought online.

It was quite a bit of fun and works very well.

The four inch square kit, uses two, Philips DF96, miniature, low consumption valves, that were developed in the mid fifties for portable military equipment. One is used as an rf amp / isolator to stop oscillation being radiated through the antenna, and the other as the regen valve and audio pre-amp. It then has a small IC amp to enable loudspeaker operation. This set is the most stable regenerative set I ever built out of several FET designs I made about thirty years ago. This one makes for very pleasant, practical listening, unlike the others I built, and uses very small amounts of power to do so. It is really easy to control the regeneration. The RF coil tuning coil and the feedback coil are two tiny 180uH inductors mounted side by side on the pcb. They actually look like half watt resistors, even down to the stripes that denote inductance value. You can vary the coupling between the coils by squeezing them nearer one another or pushing them apart. The instructions suggested mounting them on 5mm leads, so that this can be done easily.

The only fault is that it has a tendency to detect mains hum, unless the regen control is turned up and I am not certain that the directly heated cathode of the df96 isn't the cause of this. I am using a 9v wall wart power supply. The set needs an HT voltage of 12v (yes - that IS TWELVE volts) but it draws so little current that the 9V power supply is still working at its 'off load' current of 12.1 volts. When I first powered it up with a 12 volt wall wart, I found it was actually running at 16volts with the radio connected, so I quickly switched off, realising that the tiny heaters were controlled by 470 ohm, series resistors, whose value was calculated on the 12 volt supply and not 16. When I connect these power supplies to other equipment such as wifi routers, they supply their proper rated voltage.

The heater voltage is 1.4 volts (1.33 measured) and design heater consumption is 25ma, so it really doesn't draw a lot of power.

I'd have liked the kit better if the designer had used a third valve to produce the required audio output for a 3 inch speaker instead of an IC. That spoils the retro look rather. I'm not sure if the DF96 would be suitable as an audio output amp. I might buy another df96, and lash it up on some perf board with an audio transformer, or perhaps a small mains transformer with the primary in series with the valve anode and the low voltage secondary winding connected to the speaker. I suppose I could research some of the mid fifties portable radio circuit line ups to get a more suitable valve. If this works out, I would make a whole new valve only regen set, using the same front end design employed in the kit.

By the way - the computer phone/microphone headset shown in the picture has been dispensed with. I went into the junk box and dismantled an old clock radio to purloin its speaker. Ten feet of antenna can easily produce good loudspeaker volume on the Medium Wave Band.

[kalee20]

Good review.

The directly heated valves are mega sensitive to hum or noise on the filament supply - after all, it is effectively directly in series with the input. So yes, ripple from your wall wart could give mains hum.

Why not run it from a 12V battery?

As for binning the IC output stage - that's the right thing to do, these little sand-based components make excellent hardcore for concrete driveways (don't have the pins facing upwards though). But the DF96 won't do much in headphones with 12V, much less a loudspeaker. You might try a DL96 instead, connected as you say. It could be disappointing, but if you can rustle up 45V you should get results then (at least one commercial receiver used 45V with a loudspeaker output).

[neon indicator]

Yes. I thought DL96.

and a vibrator to get 90V from 12V.
Or 5 x PP3 batteries for 45V HT and one D cell for heaters.

Pity you need 10 valves to run the heaters from 12V

[JHGibson]

The only audio "Power" output valve running with 12 volts anode voltage, is the 12K5. Even then the output power is only about 35 mW, barely enough for headphone use. Two of them in pushpull could generate 100 mW, just enough to run a large sensitive loudspeaker.
John.

[Tony1951]

Thanks for the advice.

On the 'hum' thing.... I set aside the wall wart power supplies and connected an ATX, computer power supply which has a very well stabilised 12 volt output. It is a bit big, being capable of about twenty amps, but the result is loss of all hum and crackling. The power supply now dwarfs the little wireless.

Thanks for the advice on the output valve and useful HT voltage. Maybe I can get that kind of voltage with a bunch of diodes. I saw some voltage multiplying circuits somewhere to follow transformers.

[kalee20]

If you are going to use mains power, why not use mains valves? Lots of them about, no hum problems with the heater, tough as old boots - and lifetime in millions of hours not thousands!

I would suggest EF80's in place of DF96's.

In the meantime, if you want to leave your working front-end as-is, and get an audio amp up and running, try an ECL80 (you may get useable volume with a few tens of volts HT, and it has a triode section you can use as a pre-amp). There are loads of them about. Only awkward thing is it has a common cathode, which can make biasing fun.

[Tony1951]

Thanks kalee20.

I have had great help here since I joined. I am pretty much a complete beginner with valve technology. I built a small, single valve linear amp about 30 years ago to follow a double side band qrp rig I had made from a kit. I think I used a 6l6 or a 6v6 for that. I can't remember. I obtained the voltages for that from an old FT-200 valve transceiver I had. Apart from that, I have only ever been an operator of valve equipment. My favourite current transceiver is my TS530 which has a 12by7a driver and a couple of 6146Bs. I have neutralised them, and mostly leave well alone and just operate the thing.

I will look up the pin outs of the EF80 and try to figure out how I could use both its sections. The 6l6 might be another alternative, if I could make a power supply. I have a circuit for that already as an audio amp. On the face of it, it would be far too powerful for a simple radio, but maybe it would run at lower B+ voltage than 250v.....

Looking at the circuit in this little regen set, the df96 pentodes are operating as tetrodes, because the third grid is not connected to anything at all.

[AlanBeckett]

G3 is internally connected to the Filament.
http://www.r-type.org/pdfs/df96.pdf
Alan

[Herald1360]

Originally Posted by Tony1951

I will look up the pin outs of the EF80 and try to figure out how I could use both its sections.

EF80 is a simple pentode not a dual valve. It's B9A so you couldn't sub it physically in the PCB. For some info on using itin a regen, try a search on BOP Ether Ranger......

[kalee20]

The 6L6 or 6V6 may work at low voltages. I you have some, you might as well try! But they'll look a bit out of scale alongside your DF96's. On the other hand, the ECL80 will look about right. (It has a B9A base by the way). And it has the triode section.

Here's what I would do - bin the LM386 and make your audio amplifier with a B9A triode-pentode. You can have fun seeing what volume you get at different HT voltages, and optimising the bias and the load at each voltage. When you've got something working, stop and enjoy for a few weeks until you get bored. Then, get a couple of similar-sized pentodes (like the EF80), and rebuild the front-end - as Herald 1360 has pointed out, these won't fit the B7G valveholders so you might as well duplicate the circuit on a breadboard and a fresh set of components, putting aside your DF96 front end. After you have the new layout working, coupled up to your valve audio stage, you then recover the LM386 from the bin, put it back into the kit radio, connect to a 12V battery - and you have two radios!

The EF80 does work very well as a regenerative radio - although I never tried it at really low voltages.

[ParcGwyn]

Originally Posted by kalee20

The EF80 does work very well as a regenerative radio - although I never tried it at really low voltages.

Has any one tried the EF91 at low voltage? a few years ago (1981) as an apprentice I built a two valve loudspeaker radio using two EF91s based on a 1950s EF50 design from Wireless World. I think I used about 150V HT, but I would be interested to see how a EF91 performs at 12V in a regen design.

The kit looks nice by the way.

Dave

[Tony1951]

Well... Am I making some mistakes here!!

Thanks for the corrections about the internally connected grid in the df96 and my misrepresenting the other valve as a pentode/triode... LOL. I have been scanning lots of info on valves and am obviously getting confused. I am a complete beginner with this - and it shows... )

Thanks also for the advice about building a totally new regen radio.

I had thought of suggesting to the kit seller that he make a three valve one, but I can see why he doesn't do it. He markets the kit I bought, as having safe voltages. I can imagine some youngsters will maybe buy them (I would have, if I'd seen them as a youngster). So given that he is committed to a twelve volt HT line, there is no way the audio would drive a speaker with a third valve. I've been looking at the valve line ups in the portables of the late fifties and sixties and they ran HT at 90 volts from a battery. The hybrid car radios of the day ran these DF valves in the early stages and semiconductor outputs to drive the speakers, I think. They wouldn't have done that if they could have got decent volume out of a valve with a B+ of 12 volts.

So - as you guys have said - a new line up should be tried. Then the question is how to get the required voltages. I can afford to mull this over a week or two, having just spent £40 on the tiny kit.

[ParcGwyn]

This site may be of interest: http://cool386.tripod.com/12vsuperhet/12vsuperhet.html
This shows a complete MW superhet driving a loudspeaker with a supply of just 12V and it uses standard valves.

Dave

[Tony1951]

Originally Posted by ParcGwyn

This site may be of interest: http://cool386.tripod.com/12vsuperhet/12vsuperhet.html
This shows a complete MW superhet driving a loudspeaker with a supply of just 12V and it uses standard valves.

Dave

That was an interesting read Dave. Thanks. He still only got a small amount of audio though, but it is interesting that he could make the idea work so well. I also admire his mechanical construction. Very neat.

For the moment, I am still thinking about TRF rather than superhet designs.... Less to worry about there and as my little kit has shown, they can work pretty well. I must say, the design of my current project works much better than the old HAC FET designs I tried back in the late 70s / early 80s. On the other hand, I was using those on short wave and not on the medium wave band, so they had to be in oscillation all the time. As I recall, they were extremely sensitive to hand capacitance and antenna loading, and with an outside antenna, they would retune themselves as the wind blew the wire around. Hopeless really from a practical point of view when trying to copy SSB and CW. Maybe to evaluate the current design which uses control grid bias to determine the gain of the oscillator stage, I should re coil it for the 1.6MHZ - 6MHZ band and try that for size. Then we aren't comparing apples and oranges so much as now. The other feature of this kit circuit is that the first valve is an un-tuned RF amp and isolates the tuned circuit and oscillator from the vagaries of the antenna.

On the HT power thing, I went through the Christmas tree decorations box and found a couple of those AC/ AC transformers for the cheapo Chinese fairy lights I buy every year. I have a 24volt AC one which I could easily step up to 96volts with a few diodes and capacitors. I'm assuming I would use electrolytic types of a about 200uF with suitable voltage rating. The individual ones in the diode ladder wouldn't need to be high voltage types because they are only dealing with part of the ultimate voltage. Maybe 35v would be ok for all but the last one?

[Tony1951]

Originally Posted by Herald1360

EF80 is a simple pentode not a dual valve. It's B9A so you couldn't sub it physically in the PCB. For some info on using itin a regen, try a search on BOP Ether Ranger......

I meant ECL80 as suggested by Kalee20. I'm glad you pointed out my mistake though becuase it would be pretty disappointing to order the wrong valve...

:eek:

I couldn't find much on the BOP Ether Ranger except references to the print version of the Boys Own Paper. Still, I have loads of ways forward being suggested here and it makes for some interesting planning and scheming.

[kalee20]

Originally Posted by Tony1951

The hybrid car radios of the day ran these DF valves in the early stages and semiconductor outputs to drive the speakers, I think.

Not quite - they ran ECH83, EF98, and a couple of others in early stages (with transistor output stage as you say). These valves had indirectly heated cathodes and were like normal mains valves (apart from being optimised for 12V of HT).

One problem these days with battery valves - anything higher than 60V is considered a hazardous voltage so needs extra protection to avoid falling foul of low-voltage directive. So marketing similar kits to yesteryear, with 90V of HT (or ten PP3 batteries) is a no-no.

If you are making do with things to hand, and working on a shoestring, then yes you should be able to get some tens of volts using the transformers and voltage multipliers you describe. The capacitor and diode ratings should be OK, but post a circuit up for us to vet if you like.

You are right in thinking you can have a lot of fun with TRF radios, and yes they can work very well, so forget about superhets for the time being!

[Tony1951]

Originally Posted by kalee20

If you are making do with things to hand, and working on a shoestring, then yes you should be able to get some tens of volts using the transformers and voltage multipliers you describe. The capacitor and diode ratings should be OK, but post a circuit up for us to vet if you like.

You are right in thinking you can have a lot of fun with TRF radios, and yes they can work very well, so forget about superhets for the time being!

I did once build an AM / SSB, twin IF superhet, way back. I think the circuits came from a guy called Penfold. There was a series of little paperback books by this guy on electronics and radio projects. It was modular, with a 1.6mhz first IF and 455khz one with a demodulator at the end. I etched the boards and layed out the track design myself from circuit diagrams. The thing is - in the end, it worked ok, but it had cost me a fair bit and was far worse than radios I could have bought for half the money.

Here is a first idea of a possible circuit for a power supply to produce HT of 96v, 24v and a heater supply of 6.8v. I expect the capacitors in the diode ladder ought to be about 35v electrolytics or about 100uF, but I'm guessing there. I expect ordinary diodes like 1n4001 would likely do, but I would ask the parts guy about that at the electronic bits emporium. There is quite a good little old parts shop about 8 miles from here at Cullercoats by the seaside.

http://lh5.ggpht.com/_mxD813fC848/TK...0/DSCN1974.JPG


Sorry about the terrible diagrams. I had to edit out errors with bits of paper dropped onto lines that shouldn't be there.

[jimmc101]

Something not quite right there, the voltage multiplier appears to be in series with the bridge and there is a spurious wire from the the first diode /cap junct to the other side of the transformer.

There is a proven design on Paul's website here... http://www.vintage-radio.com/project...y-set-psu.html

Jim

[Herald1360]

Something a bit fundamental, too. For a quadrupler to give around 96V, the peak voltage of the ac in needs to be 24V, so you need a transformer with about 17V secondary, plus a bit to compensate for diode losses, say 18V.

Oh, and that 6.8V regulator would get a bit hot with nearly 30V across it

[Herald1360]

Originally Posted by Tony1951

I couldn't find much on the BOP Ether Ranger except references to the print version of the Boys Own Paper.

If you would like a scan of the original article, I've got a pdf, but it's rather too big for the forum. PM me your email address if you're interested.