Geordie McBoyne Wrote:As one would expect from Murphy even the most basic record player is properly designed.
Hmm. I'm not sure about that.
Refer to the BVWS Bulletin, Spring 2019 - there is an in-depth article on the Murphy A40C radio by Colin McLaughlin (RIP). It's an expensive, pre-war set, bristling with features. Colin's thorough work uncovered that Murphy operated the rectifier valves at 10V above their maximum voltage ratings. And that wasn't the only bit of iffy design.
Having established that Murphy designers were happy to be naughty when it suited them, let's see if this tradition persisted after the war, into the A316G era, with reference to the circuit diagram that Geordie helpfully put up.
First, we see that the double-wound mains transformer has half-wave HT rectification, when it would have been so easy to have used full-wave (either a bridge metal rectifier, or CT secondary) which would not subject the transformer to DC unbalance. It would give the reservoir capacitor an easier time, and result in less ripple. It might even negate need for the hum-cancelling output transformer.
Next, let's look at the amplifier itself. We see negative feedback - good! But we see the tone control INSIDE the feedback loop, so while the tone control is twiddled to give treble cut, the NFB fights to straighten it out.
We see that R1 in series with the volume control throws away 39% of the cartridge output voltage. How much nicer would the quality have been, if the full cartridge output had been used, but the gain sacrificed in some other way, NFB to the cathode of the triode perhaps. The extra loading of the cartridge (currently 830kΩ) could have been recovered by use of a 1MΩ potentiometer rather than 500kΩ.
We note that the NFB is taken to the earthy end of R1 in any case. So, the degree of NFB is dependent on volume control position - minimised when the VC is set for full volume, when the distortion-reduction effects might be most in demand with the large signal, and maximised when the VC is at zero where you get super-linear performance when the signal is guaranteed to be zero! So you can have good quality OR plenty of volume, but not both. Maybe it didn't matter to the target customers, they weren't worried about quality as long as it was loud...
As the value of C4 reduces NFB at low frequencies anyway, we note that there is a bass boost effect at lower volumes which gets less at the higher volume settings (this must be the audiomatic bass boost described in the brochure Peter Scott put up). We have already seen that the treble cut has to work harder the more feedback there is, so again, there's a volume-dependent adjustment to the frequency response at the top end as well as the bottom.
I'll be the first to admit that the circuit worked, worked well, doubtless giving much pleasure to customers, sales to keep Murphy shareholders happy, and employment to their workers. But let's not pretend that it is free from shortcomings that could have been avoided at practically zero cost to production.
This isn't complete, there'll be other observations, feel free to add to, or argue against, anyone!
...
