FM stereo test signal generator

In 2009, somebody gave me some Grundig “Decoder 6” stereo decoders and three radios that contain one of these decoders. I was able to find some aligment instructions. As it turns out, most of the aligments can be made using just a precise AF generator and a scope. But to set channel separation, you need a stereo encoder. So I started to look for stereo encoder designs. I looked for a traditional analogue design.

Building a complete stero transmitter would be easier than I thought. I discovered a number of IC types that contain a complete stereo encoder and even a PLL-controlled FM transmitter. These chips are used in large quantities in the tiny mini FM transmitters you can buy to play the music from your MP3-player on your car stereo.

I found out that ELV are offering an FM signal generator kit, “SUP1”, based upon the BH1415 chip from Rohm. This ic contains a stereo multiplexer with two filter pre-amps. The 19 and 38 kHz signals are derived from a 7.6 MHz crystal oscillator. The crystal oscillator is also the reference for the PLL frequency synthesizer that has a serial input to set the frequency multiplication factor. The frequency can be set anywhere from 87 to 108 MHz.

The SUP1 test generator has a microcontroller to set the frequency and switch from stereo to mono mode and an LCD display. I decided to order one.

My kit arrived from Germany by mail a few days later. It was easy to assemble. ELV have already soldered all the SMD components to the PCB, leaving only the wired components and switches and connectors for me to solder.

I checked the circuit, connected a battery and threw the switch. The display came up as described in the instruction booklet. I connected my scope and saw the RF output signal. On an FM receiver I got a carrier, right on 87.5 MHz. I set the frequency to 108 MHz and that worked fine, too. Then I connected an AF generator to one of the inputs and heard the tone through the speaker. I set the mode button to stereo, but now the sound was slightly distorted. On my scope, I saw that the stereo encoder can only deliver about 120 mV top before it starts to clip.

Now I wanted to hear some music. The guys from ELV are warning that it is illegal to connect the generator to an antenna. So I used a balun to connect the generator directly to the antenna input of my Philips B8X44A FM stereo receiver.

Now something odd happened. When I tuned to the generator frequency, I only heard white noise. But when I pulled the balun from the antenna input, I got strong reception of the carrier. Then I connected an MP3 player to the input and heard the music. After some experimentation, I concluded that the signal from the test generator was just too strong for the radio. The strong signal stops the local oscillator of the FM tuner, so that no frequency conversion takes place and only the noise of the IF amplifier can be heard. I tried this on another Philips receiver that had a similar FM input stage, my A5X83A AM/FM tuner and got the same result: noise if the generator is connected to the antenna input, and reception when it was disconnected. So I needed to use an attenuator between the test generator and the radio.

I tried the test generator on a few FM stereo radios to judge its performance. I found the sound level is somewhat weak. In stereo mode, distortion and hissing noises can be heard. Channel separation is not impressive. If you connect a tone to one channel and listen to the other one, you hear a weak tone with distorsion and noise.

Well, after all, this is only a simple stereo encoder. Doing some more research, I found that people who are fussy about putting a decent FM stereo signal on the air, such as the Pira.CZ pirate radio site, hate the BH1415's performance and prefere a more complex DYI stereo encoder.

So it looks like the channel separation of this test generator is too poor to be useful to align my stereo decoders. But it can be used as a simple FM test transmitter. Anyway, nice try.