The transmitter works! Somehow youtube is not processing my videos right now.

The dynamic range is a problem.
The circuit has drift.
The coils actually work, with the slugs nearly fully inserted.
Adjustment of the oscillator capacitor is difficult.
The preamplifier is unproblematic.

The output transistor heats up! Actually it works only at a very narrow point (I made R7 adjustable). So I guess the 2SC1740 does not have so much gain for this circuit. The C13 also needs adjustment within a relatively small area. But not as difficult as the oscillator capacitor.

The 2SC1740 does not seem to be working as well as 2n3904 maybe would.
Having an adjustable bias resistor for the preamp is also helpful, since otherwise there is a lot distortion.

OK the video has processed: http://www.youtube.com/watch?v=YDi4XlXm8Vw

It is a cheap receiver of course. In the video I use different volume settings, both on my PC, and later I change the radio set volume. Not all of the distortion is produced by the transmitter.

I can receive the signal at several frequencies, but except one they are not are useful, the others are all weak and full of distortion.

 

Maybe I could duplicate the input stage, and alternatively blank out one. Then somehow filter it, and mix with the pilot tone.

You need to learn about the FM stereo signal. The first part of the article I posted explains it.
Basically you need L+R mono audio and you need L-R modulating with AM a 38kHz suppressed carrier so that the L-R sidebands are from 23kHz to 53kHz.
Then you add L+R with L-R to get 2L and subtract L-R from L+R to get 2R.
Then add half the frequency with exact phase of the 38kHz carrier to make the 19kHz stereo pilot tone. Filtering is also needed.

 

The transmitter works!

Good.

The dynamic range is a problem.

I heard mains hum on the video so maybe the input cable is not a shielded audio cable.

The circuit has drift.

It is much better if the pcb is inside a metal box (but spaced away from the metal) connected to 0V.

The coils actually work, with the slugs nearly fully inserted.

What frequency is it at? Nobody else out of the hundreds on the internet use ferrite cores on the coils.

Adjustment of the oscillator capacitor is difficult.

A metal screwdriver changes the frequency. Use a plastic screwdriver instead.

The output transistor heats up!

Yes, its hFE is high and it has an extremely simple biasing from R8.

It is a cheap receiver of course. In the video I use different volume settings, both on my PC, and later I change the radio set volume. Not all of the distortion is produced by the transmitter.

If the input signal to the transmitter is too high then its FM modulation swings its carrier frequency too far. Since the radio IF bandwidth is set for the level from normal radio stations then anything louder than a radio station feeding the radio causes severe distortion. Turn down the level feeding the transmitter so the radio is not louder than an FM radio station.

I can receive the signal at several frequencies, but except one they are not are useful, the others are all weak and full of distortion.

Cheap radios are easily overloaded by a strong local station causing the station to appear at many points across the entire dial. My cheap Sony Walkman radio fixes overloading with a "local" , "distant" switch. In the local position the antenna input is attenuated so there is no overloading but also no distant stations. In the distant position the dial is full of a strong station or two at many points on the dial so again there are no distant stations.

The transmitter is a very strong local signal. My high quality home stereo and car radio are never overloaded by anything. But they are also very sensitive so they pickup my transmitter signal when it is far away.