Valve cassette decks, again...
 

I read in the past some threads about valve cassette decks. I have been working sporadically on a valve playback circuit. It uses a jfet input, but otherwise 3 valves for most of the gain and E.Q. I could use a step up transformer and or try lots of parallel valves for the low noise, but the jfet input buffer is very convenient, and the whole circuit is subjectively quiet. Ultimately for me this is about seeing how the tape sounds with a decent valve amp rather than all out principles.

I am feeling quite pleased with it at the moment, as my most recent round of tweaking has resulted in a very satisfying sound quality. In my investigations and researches I concluded that to get Dolby B to work properly the higher frequency performance is absolutely critical. Being in complete control of the design, which to be honest is a bit seat of the pants, has ended up with decent Dolby friendliness.

I am running it with a wireless world decoder at present. I have almost got a Dolby B decoder running, using valves for the audio path the "jfet" and transistors for the side-chain, but circumstances have halted the development of that. Now the deck is running pretty well I feel more inspired to pick it up again.

Re: Valve cassette decks, again...
 

That looks very impressive! Consider me interested.

Re: Valve cassette decks, again...
 

Yes for Dolby B the overall response, theoretically up to 20kHz, needs to be close to flat. The tape recording process will be the weak link, as it always was, especially with cassettes. Transistors and valves shouldn't be the issue although of course the insertion of transformers with valves might be a cause of non linearity. Interestingly, the later Dolby C while more touchy in other respects is less sensitive to high frequency errors as it is deliberately only responsive below 10 kHz, a feature they probably should have included in the original Dolby B.

Again I suspect that if you do it well, the valves will make little difference. The weak link will be the tape recording and playback process itself, especially its alignment when Dolby is involved. Type II or even Type IV tapes make a difference.

I tend to go the opposite way with tape machines, replacing older valve circuits or inferior transistor components such as noisy Germanium trannies with later lower noise designs.

I suspect a well designed older valve reel machine ensured the tape saturated before the valve circuitry did, so the latter was effectively transparent. That was also the aim with transistor circuits. It was always something of a compromise especially with 2 head cassette machines where you needed a relatively high impedance head for low noise playback but relatively low impedance for recording to avoid the record amp clipping.

Tim

Re: Valve cassette decks, again...
 

My feeling is that good sound can be achieved with either solid state or valve. It is more an issue of topology. However I think valves have a head start due to the different nature of their distortion. If they really were not appealing to mankind on some level they would have gone away like steam cars and horses and carts.
The avoidance of feedback around multiple stages is key in my opinion to achieving the most satisfying sound. Putting feedback around more than one valve stage will make the valve circuit sound more like the typically "transistor" sound.