Civilian Wartime Radio - Further Notes

[ppppenguin]

Alistair cited this article which explains the whole horrible business of asymmetric modulation: https://www.thebdr.net/support/xmtr/maint-ammod.pdf

Anything to squeeze out a bit more apparent loudness and coverage without increasing the electricity bill or exceeding the licenced maximum power.

[Alistair D]

My last sentence should have read;

Secondly; The receiver's AGC is confused into thinking the signal is smaller than it really is so it may not turn the gain down enough.  

Sorry

Al

[Terry]

The receiver's AGC is confused into thinking the signal is smaller than it really is so it may not turn the gain down enough.  

But is it?

Surely the AGC detector will respond to the higher peaks of the signal and attempt to reduce the gain as the signal will appear stronger? [1] The actual reduction will depend on the overal amount of limiting applied and the time constant of the AGC loop.

[1] Indeed, because of the higher peak levels, it can be argued that, on peaks anyway, the signal strength is higher.

[Alistair D]

Terry, having thought about it again I realise that statement is wrong. I had confused myself about the AGC rectifier rectifying the -ve going part of the I.F. output waveform. 

My excuse, galloping senility.

Al

[Sparks]

This is how it looked on completion.  Various component values were subsituted because I hadn't got any of the original values.     It is all tidied up now and working very well indeed.     The modulation on the local station is still slightly scratchy, but not all that bad.   All the rest are very good.    The Westector is now back in again, but the germanium diode worked fine when it was there..
Bob
   

[Terry]

 Various component values were subsituted because I hadn't got any of the original values.

Not surprising in many cases as the original values are no longer available! At the time these sets were made, the tolerances of capacitors were such that, except in unusual circumstances, the values of the waxed paper capacitors followed a 1, 2, 5, 10 etc, series. (From memory, I think some manufacturers used 2.5 rather than 2, but it makes no difference.)

In the 60s, the polyester types that replaced them could be manufactured to much higher tolerances and started to be introduced with values in the E6 series, thus 1.0, 1.5, 2.2, 3.3, 4.7, 6.8, 10.0 - there being 6 values per decade following a logarithmic series.

Thus you had to replace all of the values ending in 5 with values ending 47. If there had been any 0.2µ, 0.02µ or 0.02µ capacitors you would have had to use 0.22µ, 0.022µ and 0.0022µ (or 2.2nF) values. In fact you did have to use a 22µ electrolytic for the same reason. Unless anybody redraws an old circuit from scratch, I doubt if anybody would bother to update it.

If you want to understand why resistor values, in particular, have such an apparently strange sequence of values, this chart will be of use to you:



Click on the image for a larger view.

You will immediately see that all the values neatly butt up against their neighbours (or very, very slightly overlap) at the extremes of their tolerance range

[Sparks]

Thanks, I had wondered.
In the circuit shown above, one is incorrect. The cathode bypass of V3 was originally 25uF. I had made a note on the diagram to put a 22uF one in, but when I wired it up, I inadvertantly left it off altogether and when I tried it, I could hardly hear anything, even on maximum volume. I had it upside down and switched on, and the only capacitor to hand, lying on the table, was a 0.1uF. I held it across the cathode resistors and immediately, the voulme shot up to normal. I never even tried the 22uF that I had intended to fit, but soldered the 0.1 in! I may try a 22uF eventually, but at the moment do not wish to mess about with it because I am giving a talk about it on saturday to the local model shipbuilding society, and then on the following Monday to a Royal Air Force welfare society that saw it on Facebook and for some reason would like to hear about it, and see it working.
Bob

[Crackle]

Excellent! I hope your talks go well Bob

Mike

[ppppenguin]

I don't know the value of the cathode resistor but if it's 1K, the resulting frequency response will be 3dB down at about 1.6kHz. If the resistor is 500R then 3.2kHz. In other words therewill be very little bass with a 100n cap.

[Murphyv310]

I think on those sets the detector will not properly work if the cathode isn't bypassed by a value near its original, the biasing for the detector comes from the cathode of the 6V6 so too much AC content will cause problems.