Avantic BM611 Stereo Decoder

[PJL]

Although it is detecting the pilot tone, the stereo separation is very poor. I assume this is due to attenuation of the upper multiplex channel. It seems improbable that this will be IF bandwidth so my thinking is loss is in the FM discriminator or audio circuit. The 2022 siggen is only specified to 25KHz external modulation so I have some problems with test equipment.

[Nuvistor]

You can use a standard sig gen to test the bandwidth of the IF and det but the gen needs to be accurate.
It's a Foster Seeley det and I forget if it as a DC output like a ratio det, but you could used a diode probe to det the output.
Monitor the output while swinging the gen across the IF range, then calibrate the 3db points.

For linearity the det will require a much wider bandwidth than the IF, some better det stated around 1Mhz but it will work well with less.

[PJL]

I did align it using the 2022 to get a fairly flat +/-99KHz using 1kHz modulation so I am not convinced it is IF bandwidth that is the issue. What I would like to do now is test how the FM detector and audio behave with high modulation frequencies.

I could do a better check of the IF bandwidth as I have a sort of working polyskop in the garage but it is kind of heavy!

[G8HQP Dave]

200kHz IF bandwidth is not really enough for good stereo. You need more like 250-280kHz. FM uses the sidebands at odd harmonics of the baseband to carry the signal, and those at even harmonics to maintain a constant RF amplitude. Stereo MPX has signals up to 53kHz so even 318kHz IF bandwidth only just allows for the third harmonic at +/- 159kHz - and that ignores phase issues.

Looking at it another way, if there is much stereo information in the original signal then this means lots around 38kHz. You want this to get through undamaged so you need low distortion and accurate phase and amplitude. To get just the 3rd harmonic sideband through OK means an IF bandwidth significantly greater than 6x38kHz = 228kHz.

Another issue is that you do not necessarily need a flat IF passband, but good phase response. This means a rounded passband, quite different from AM; Bessel rather than Butterworth.

I suspect that some of the earlier FM receiver designers did not realise this and got poor stereo from their add-on decoders as a result.

[PJL]

I have it back on the scope/siggen and it is showing some odd behaviour (hysterisis) through the cathode follower so I am going to try taking it off the detector but it will be marginal without a buffer. The bandwidth is probably a bit tight showing a pretty linear response over +/-99KHz when the signal is just triggering the tuning indicator but take the signal down 10dB it is clearly rolling off at the edges. So next steps are:
1. Improve the signal path from detector to decoder including build a high input impedance buffer.
2. Drag out the Polyskop and have another go at alignment.

[PJL]

I have proper stereo now after temporarily re-routing the detector direct to the cathode follower to give a bit more output and tweaking the IF alignment. I have over done the alignment and now have multiple peaks so need to have another go with the Polyskop. Next job is to build the JFET buffer and voltage doubler.

[PJL]

I have had another go at alignment using the 2022 but this time I moved the centre frequency about to cover a wider range. After the alignment I checked the demodulator output and I now get a good straight line output over +/-200kHz.

Whilst that proves the FM detection range, the other factor will be the demodulators ability to handle the high frequency multiplexed audio. In the drawing of the Foster-Seeley demodulator below there is a 40pF capacitor between the diode cathodes which I assume is RF filtering. I haven't a clue how to calculate the impact of this capacitor and don't have the test equipment to generate a 53KHz FM modulated signal on a 10.7MHz carrier to measure it.

The only solution would be to try a few values and see what impact it has on the stereo quality?

Meanwhile, I am just testing my simple J-FET buffer which will be needed to deliver the audio multiplexed signal to the decoder without losses.

[Chris55000]

Hi!

That tuner is a BRC/Thorn Diagram , so almost ceriainly it's their manufacture, and so I suspect there's an equivalent model in the Ferguson/HMV/Marconi/Ultra ranges that corresponds to that circuit - I'll see if I can lay my hands on any further information, circuit values or a Thorn MPX adaptor!

Chris Williams

[PJL]

The buffer is in and working but I now need to add a volume control to the decoder as the signal is too high! I also replaced a few other parts that had been snipped & tacked by a previous owner presumably during fault diagnosis, and changed the IF filter capacitor from 40pF to 18pF.

I have been using a valve amplifier but I need to wire it to a solid state setup so I can listen on headphones only. I will then have yet another play with the IF bandwidth to see whether I can get away with a tighter setup.

[PJL]

Made the changes and have it on headphones. It very readily locks to the 19kHz pilot tone. Separation seems very good although I have no means of measuring it. There is a very slight tendency for sibilance on some pop music tracks appearing in the extremes of the channels. Some hiss is evident on weaker stereo stations but it seems far better than previous receivers I have owned.

There is quite a wide tuning range where the signal largely remains in tune and I do wonder if the extra IF bandwidth is really a benefit or if it causes problem with close-by channels. I suspect narrowing the bandwidth will mostly impact stereo reception on distant stations. I will have one more go at the IF alignment but will try reducing the IF filter capacitor first.

[PJL]

Decided to leave as is. I started looking at the waveform out of the popular music stations and the incredibly heavy compression and clipping is astounding. Most of the time it is rapidly swinging from one legal bandwidth limit to the other displaying a very decorative band across the centre of the scope. I checked and I can create a larger swing with my siggen so its not the receiver. Radio 3 and 4 behave perfectly.

[G8HQP Dave]

Pop music stations gave up on sound quality years ago. That may be why the only people who listen to them are builders, decorators and central heating 'engineers'.