This unit transmits audio to any AM receiver. Range is dependent on the aerial used.
What is a Phono Oscillator?
The "Phono Oscillator" goes back to the 1930's and was a popular radio accessory in the U.S. It came about as a convenient way to connect a turntable to a radio that did not have pick up terminals, as well as just having the convenience of a wireless connection.
A phono oscillator is simply a low power AM transmitter, operating in the broadcast band. It takes the audio signal from the phonograph pickup, and uses it to modulate the RF from a medium wave oscillator. Hence, the name is drived from "phonograph oscillator."
These devices also were also the basis of the popular "home broadcaster" devices which became popular later on. Instead of a gramophone pickup, a crystal microphone was used. More eleborate units included both microphone and pickup connections with a simple mixer.
The units were sometimes included inside the actual turntable, or particularly in the "home broadcaster" guise, as a stand alone unit.
In Australia and other countries, phono oscillators went unknown as conservative goverments considered anything but a licenced transmitter illegal, regardless of how low power it may produce. So called "home broadcasters" in this country were a poor substitute, being merely a carbon microphone with battery and transformer, intended to feed the pickup terminals of a radio located in another room, via a length of cable.
Most simple designs use a pentagrid valve. An oscillator circuit, similar to that used in a receiver is used. But instead of the oscillation being modulated by an incoming RF signal and converted to the IF, an audio signal is used to modulate the electron stream in the valve, and the modulated RF is then taken from the plate.
In fact, with a suitable switching arrangement, it would be possible to convert most valve AM superhet receivers into a transmitter.
Typical valves used are 6A7, 6A8, 6SA7, and 6BE6. A short aerial is lightly coupled to the output circuit and it is intended that the transmission range just covers the house.
Home broadcaster units were more powerful
versions of phono oscillators, using separate oscillators and modulators,
and based around medium power audio output valves.
Typical American kit designs would use valves like 50C5 for the oscillator and modulator with a 12AX7 for the audio preamp. It can be imagined that fed into a decent outdoor aerial, such a device could transmit for a few kilometres.
In the present day, a phono oscillator finds use in being able to convert the output of an FM tuner, iPod, CD player, or any other audio source, to AM. Once AM transmissions are switched off, these will be an essential device to hear DAB broadcasts via an antique radio.
My Phono Oscillator.
I set about designing a simple low power pentagrid valve circuit.
For the valve, I decided on a 6CS6. This is of very similar construction to a 6BE6, and as a frequency converter in superhet receivers can usually be used as an equivalent substitute. The 6CS6 found most use in TV sync separators with noise inverter circuits. Also, AWA used it as an FM detector in their P1 and P2 chassis TV sets. It is therefore a common TV valve in Australia.
To use parts at hand, and to keep construction simple, I decided on a plugpack power supply, and build everything into a zinc diecast box.
Circuit of the Phono Oscillator. Coil connections are looking underneath at the pins. 6CS6 must be used with this circuit.
A 10V 1A plugpack feeds the unit. This 10VAC supply is stepped up to about 180V by means of a 240-12.6V 150mA transformer (type 2851) in reverse. Rectified and filtered, it provides 140V at about 10mA. The valve heaters are fed via a dropping resistor to provide 6.3V at 600mA. Of course, had a 12V plugpack been used, the heaters could have been wired in series with less current drain. But, 10V 1A plugpacks were once commonly available with dial up modems, and hence I had some looking for a use. The connection of the dropper resistor might look strange, having the resistive element apparently shorted out. A closer look will reveal that, actually either side of the tapping is in parallel. This way, heat dissipation is spread over all the resistor, rather than just the part between the slider and one terminal. The slider is set so that the valves are fed with 6.3V.
I used 100uF filter capacitors which are overkill, but they were to hand. Anything down to about 10uF would do. Too low, and hum will become evident. The relatively high resistance of the 240V secondary winding limits the rectifier current when the supply is first switched on. The rectifier is a W04 bridge rectifier.
Obviously, other power supply designs can be used, provided the valve heaters receive the correct voltage, and the B+ is around 140V.
To save winding any coils, I used a transistor radio oscillator coil. These are standard, being coded with a red core. In Australia, these have been readily available as part of a set purchased from the likes of Jaycar, or once upon a time, Dick Smith. Included are three IF transformers. With a variable capacitor and ferrite rod aerial, obtained from the same stores, one can make a transistor superhet receiver.
These oscillator coils are intended for autodyne converters, which are standard with transistor MW superhets. I first tried the coil in the intended way, feeding the 6CS6 cathode to the tapping on the main tuned circuit. It worked, but oscillation dropped out at only 75V. As we know from this circuit, it should work down to at least 12V.
So, I reconfigured the circuit and used what was meant to be the collector winding as the cathode feedback winding, and got strong reliable oscillation.
Grid leak bias is produced by the 33pF and 22K in the normal way. Note the grid leak is taken to cathode potential, and not ground, otherwise the the grid leak action would be affected by the cathode bias.
The screen grid / oscillator plate are fed in the usual way via a 5.6K bypassed with .068uF. These values are not critical, and anything up to about 10K will do as a screen resistor. Likewise, .047uF and above will do for the bypass.
I set the oscillator for 1350Kc/s as this is a clear part of the band in my area. The 33pF across the coil may have to be altered for other frequencies.
Built in a diecast box, this is the underneath view. The oscillator coil was mounted on a small PCB in the lower right corner.
For this circuit, a 6CS6 must be used. A 6BE6 was tried but output was considerably lower; about 18Vp-p versus 70Vp-p into a CRO with 10x probe. Modulation was much less sensitive and some distortion was visible. That is not to say a 6BE6 or other pentagrids will not work, but circuit alterations would be required. An interesting idea would be to adapt the circuit to use a single 6V9 triode heptode valve. Note that this valve requires a 10 pin socket.
This is simply an RF choke of 1mH in the plate circuit. It is here the RF output voltage is developed. There is about 9V p-p when measured on a CRO with a 1x probe. With a 10x probe, output is around 70Vp-p with modulation. A 470pF provides isolation so the aerial is not a shock hazard.
Ideally, the output should be tuned for efficiency. Indeed I did some experiments with a ferrite loopstick and variable condenser which confirmed this. However, so as not to have to retune with every different aerial used (due to loading effects), and that the transmission range was already satisfactory, I kept the simpler untuned circuit.
If you want the utmost range from this circuit, tuning and aerial matching will be necessary.
Normally with such circuits, the pickup signal would be fed into grid 3. This voltage on this grid controls the RF that reaches the plate. I found that around 7v p-p was required to get 100% modulation. This sort of voltage does in fact come from old magnetic pickups - of the kind that can drive an output pentode to full power, but a modern audio source would require a preamplifier.
The modulating grid is held at earth by 100K for DC, and 220pF for RF. The modulating signal is coupled by .022uF. For linear modulation, the 6CS6 is operating as an audio amplifier, and thus bias is required on the modulating grid. Conventional cathode bias provided by a 560R is used here. RF bypassing is provided by .22uF. While audio bypassing increases the modulation sensitivity, it is not by much, and not worth including the required electrolytic bypass. For those than want to try it, use a capacitor of 22uF or more.
Audio is fed into the RCA sockets as either a mono or stereo signal. The step up transformer and heater dropper are mounted above the chassis. Oscillator coil is at the left.
In order to get sufficient modulation from typical modern line level sources (around 250mV-500mV), a simple triode amplifier was needed. I used a 6AV6. The diodes are not used.
There are a multitude of valves one could use here. The gain required is not high, and in fact the cathode resistor is left unbypassed. The circuit is very typical, with a 100K plate load and 560R cathode resistor. Modulation level is controlled by a 500K pot. This value was chosen to allow for high impedance valve devices to feed the modulator.
Two 10K resistors provide isolation between channels when a stereo input is used. The modulation level is simply set so the volume is the same as other AM stations.
Upper waveform shows modulated RF output. Lower waveform is the audio fed into grid 3 (pin 7) of the 6CS6. Note the level here is 5.72Vp-p.
This did surprise me. I was expecting to only transmit across one or two rooms, but with about 1m of aerial, it could be received on a Grundig Yacht Boy receiver at the end of the backyard about 50m away. Connected to the long wire outdoor aerial, reception was clear all over my property. I have not tried yet, but I suspect it would transmit over a few other neighbouring properties as well.
I used my 12AT7 FM tuner to provide the program source. It was quite strange listening to an FM station relayed to an AM portable, but you wouldn't know it was an FM station except for the call sign.
This set up with an FM tuner feeding the phono oscillator allows FM programs to be heard on all AM receivers in my house and backyard.
A logical step would be to build the phono oscillator and FM tuner on the one chassis, making a stand alone FM to AM converter.