AM Modulation

[Variometer]

Over the years, I have built a number of small valve radio sets. A slightly annoying feature, common to all of them, is that the local AM station always sounds rather "scratchy." This is more noticeable on music, but speech doesn't sound too bad. Stations further afield sound OK. Do you think this is because of the proximity or do some more modern stations have a slightly different type of AM that works OK with transistor sets, but does not quite agree with valves? The station in question is Radio Lancashire and I am located in Penwortham.
Bob

[russell_w_b]

AM broadcast transmitters use 'AMC' - AM Companding, which alters the carrier level so that 100% modulation no longer equates to twice the voltage of the carrier (or four times the power), but somewhat less, depending on the level of modulation. I can't remember the figure for MF, but HF AMC is compressed by 6db so that the peak voltage for 100% modulation is the same as that of the unmodulated carrier and never exceeds this limit. As I remember, 100% mod on MF sites exceeds the unmodulated carrier voltage level, but not by twice the amount.

This is done to save electricity, and is reliant for 'transparency' on the listener's receiver AGC. It may be that modern sets have a faster AGC which might interpret the dynamics of the programme content differently. I don't know about 'scratchiness', though... AMC is used in conjunction with audio compression (usually 'Optimod') which gives a 'punchier' signal, but alters the dynamics of the material being broadcast.

Perhaps your comparative stations are set up differently? Just a thought...

[Mark_4CX35000]

3 dB compression for BBC nationals (Radio 5).

Not certain if a small service such as Radio Lancashire would run AMC,
as there would be no real power saving against the costs of the AMC
equipment.

It might be the settings the Optimod is operating which means music
sounds worse than voice.

[Variometer]

Thanks for replies, although I have never heard of Optimod, I am aware of different types of almost compatible AM.

Bob

[Ian - G4JQT]

Hi Bob,

What you describe is most likely to be due to the receiver not handling the much stronger signals from local stations.

All AM stations (LW, MW and SW) use various types of audio compression to keep the average modulation level up to achieve a better signal-to-noise ratio, particularly at the edge of the service area.

If this were not done, and modulation was set so that 100% was reached on the peaks of, say, orchestral music, (perhaps a symbol crash every 10 minutes) the quiet passages would be almost lost, or listeners would have to turn up the volume, then turn it down for loud passages. (OK, so we don't actually get much classical music on AM these days...)

Audio compression thus pushes down the loud and brings up the quiet.

Trimming off the highest frequencies to a max of about 4.5 kHz (but varies somewhat from station to station) also maintains modulation bandwidth within the 9kHz channel spacing.

Getting all this right is quite an art. And actually virtually impossible so as to make it sound good on a small-speaker portable, a car radio, hi-fi with lots of bass, and a nice old valve set.

FM more or less gets around these problems.

Unless R Lancs has a problem (do 855 or 1557 kHz sound the same?), I don't think the fault is at the transmitter.

BTW Bob, I used to live in Rawstorne Rd/Cop Lane, Penwortham! Small world.

Regards,

Ian

[Variometer]

OK Ian, thanks for further reply. More distant stations are all OK, so maybe I am just too close. Not far from Rawstorne Road, but at Plough end of Cop Lane.
Bob

[Mark_4CX35000]

The transmitter for Radio Lancashire is in Hoghton. Just off Riley Green
Switch Road.

There is also another transmitter. Transmitts local radio Magic 999.
Located in Longton. Just off Hall Lane.

[Boom]

What happens if you attenuate the aerial input. Does this stop or reduce it?

It could be that the first rf stage is being overloaded but I'd have though this highly unlikely with a valve set using agc.

[Variometer]

It is just as bad attenuating the input RF. This has been going on for years, but until finding this forum, I had no-one to ask. It is apparent even on one valve TRFs.
Bob

[Ian - G4JQT]

If it really is bad on all sets in the house and on neighbours' sets, it could of course be a transmitter fault. Is it the only station that is strong and/or distorted? If other stations are distorted do they come from the same transmitter site?

It's not easy to find who to inform when there is a problem - assuming the engineers don't already know.

Regards,

Ian

[Variometer]

It sounds OK on battery valve superhets, it is just on my home-made TRFs that the trouble is apparent. It is not complete distortion. Speech sounds OK, but sharp, but music sounds slightly raspy. It has been like that for at least twenty years, so I doubt it is a transmitter fault.
Bob

[Mark_4CX35000]

It might be the modulaton peaks are being clipped in one of the
receiver stages. That can make something sound crispy.

Does the same problem occur if you tune into Magic 999. Around 300 metres.

Listen to varous types of progamming thoughout the day, see if the same
problems occur.

If it was a transmitter fault, the on site transmitter monitoring and telemetry
equipment would soon have all the local Tx engineers attending.

[Skywave]

Originally Posted by Variometer

It sounds OK on battery valve superhets, it is just on my home-made TRFs that the trouble is apparent.
Bob

That makes me wonder if the filter cct. in the detector anode is not quite right and inter-mod. distortion is consequently taking place. Can you put up a cct. of one of the TRF sets so affected?

Al / Skywave.

[Variometer]

Here is the circuit and component values. I have not copied this from a book, it is from something I wrote some time ago. I have tried changing all the values, but the slight distortion always remains on the local station.
Bob
COMPONENTS REQUIRED
V1 6C4 triode valve.
Valveholder B7G type.
C1, C2 100 pF capacitors.
C3 330 pF capacitor.
VC1 variable capacitor. (see text.)
L1, L2 & L3 100 uH chokes.
L4. 4.7 Mh choke. (Note that the value of L4 this millihenry, NOT microhenry)
R1 1 Megohm resistor.
VR1 47 K miniature potentiometer.
Jack socket Mono socket 6.35 mm (1/4 inch).
Sundries Baseboard, H.T. & L.T. sockets, PCB pins etc (see text).

[Oldtimer74]

Surprised that there is no bypass capacitor across the phones which of course have to be high resistance (impedence), 0.001 mfd would be usual.
Pat

[Oldtimer74]

Further to my previous. I am surprised at the coils L1, L2 and L3 all being 100 mH. L1 and L3 should be smaller than L2 , which is the main tuning coil. A variable capacitor in the aerial instead of the 100pF might help on strong signals. I presume ion local stations you can reduce the reaction considerably. If not this narrows the bandwidth which may be the cause of your distortion. I have to say I have never used or seen your system of reaction before.
Pat

[Variometer]

A capacitor across the phones does soften the sound a bit, but also reduces the volume. I use a matching unit made from a small PBC transformer 240 V in 12-0-12 out. Have also tried it on 4,000 ohm headphones & performance pretty much the same. At the time I made the circuit, I was just starting to use RF chokes as coils (never did like winding my own). The values were just a guess, but gave good coverage. In later sets, L1 is 47uH, but the other two remain at 100uH. Only used this type of reaction circuit because small variable capacitors are becoming rare & expensive. A later and very smooth reaction unit is described lower down Homebrew, where I have the chokes inside old 1 14 inch fuse cartridges and the reaction choke actually turns around physically.
Bob

[Oldtimer74]

OK I see. RF chokes are not really good for use as tuning coils as the Q is not designed to be high. Re the reaction. If you were to change the valve to a pentode, EF91 or similar, you could then control the reaction by a potentiometer controlling the screen voltage. You only need a trimmer to set the reaction feedback, which once set is never touched again. This gives the smoothest control I know and works well on SW as well. I have a 2 valve receiver I made using Eddystone plug in coils using such a circuit.

Pat

[Variometer]

Hi Pat,
Thanks for that most useful reply. I have the book Radio For Boys, 1951 and that uses a similar method for reaction. The capacitor is fixed and reaction is controlled by a potentiometer that adjusts both anode and screen grid volts to the detector valve. I have never tried that system, but will definately give it a go as my latest circuit is using pentodes. As for the choke/coils, I really cannot fault them. I built a four valve superhet using only chokes and it performed well and the tone was OK on Radio Lancashiore as well. They don't seem to work on SW though, but I haven't tried very much as yet. I even used them for the IF transformers (1mH with a trimmer across them). My rotating coil reaction is superb, but it does take time to build as it requires bushes & spindles etc.
Bob

[Sideband]

I noticed some distortion on AM 'locals' with my Philips Superinductance receiver. That does use an anode bend detector though which requires a critical audio input to work well. It just cannot cope with the compressed audio from the 'pop' locals. It sounds fine on 5 Live, Talk Sport and R4 though. Also it seems to be OK on Continental music stations. Whether this is because they are weaker or because they don't use audio compression I don't know .



Rich.