What is it about a circuit.

[r3xj0hn570n]

Yeah, so the fundamental problem is that asking a single active device to amplify voltage doesn't work very well. BJTs have horrible distortion characteristics, FETs less horrible, valves less so again and can be linear over a small region. Nelson Pass's site has a good explanation on these different devices. Anyway, people being the clever things they are create all sorts of circuit combinations to try a mitigate these non-linearities. There is local degeneration, trading voltage gain for linearity. There is feedback which samples the output, inverts it and feeds it back into the input by all sorts of extremely clever arrangements. There are distortion cancelling arrangements where one devices distortion cancels the other. Each of these has drawbacks. Trying to isolate the 'sound' of a single component in all this complexity is fraught. Capacitors for example, good old polycarbonate caps are awesome when used as main voltage bypass, but less so when used as simple interstage coupling caps. Something to do with the dielectric polarity IIRC. Inductors saturate, so are awesome in low level positions, awful passing greater current. I haven't even touched on noise, of which there are many types.

[Geoff.R.G]

I am difficulty believing different circuits with the same parts will sound different.

Undoubtedly they sound different but you need more data, Colin suggested Mullard data books, which I have suggested you will find on line. I am too far from recent to know whether you can use the same transistor in all three basic configurations, you would need to consult the data. However I suggest that the most frequently used configurations, common collector and common emitter, are favoured for good reasons but try them and see what you think.

Common Base, Common Collector ,Common Emitter , how about Kerchov config, that makes 4.????

I am not familiar with the Kerchov (Kerchoff?) configuration Colin, would you care to enlighten me with a diagram please?

[Wonfor14]

OK I can't spell.
Current into point must equal current leaving a point, or in this case it could look like a push-pull Common Base.

See link:- [https://seca.freeforums.net/thread/325/pp-common-base][/url]

I patented this in 1978 and Magnum used it and paid me a licencing fee.
It was used in a MC gain stage, the transistor were matched on a common die.
I took it to Rotterdam Uni and described it to Prof Van-Develt and he said it could not work, I took my vero model out and showed him it worked.

Geoff if you could find away to put pic up here I would love it please.

[Wonfor14]

Yeah, so the fundamental problem is that asking a single active device to amplify voltage doesn't work very well. BJTs have horrible distortion characteristics, FETs less horrible, valves less so again and can be linear over a small region. Nelson Pass's site has a good explanation on these different devices. Anyway, people being the clever things they are create all sorts of circuit combinations to try a mitigate these non-linearities. There is local degeneration, trading voltage gain for linearity. There is feedback which samples the output, inverts it and feeds it back into the input by all sorts of extremely clever arrangements. There are distortion cancelling arrangements where one devices distortion cancels the other. Each of these has drawbacks. Trying to isolate the 'sound' of a single component in all this complexity is fraught. Capacitors for example, good old polycarbonate caps are awesome when used as main voltage bypass, but less so when used as simple interstage coupling caps. Something to do with the dielectric polarity IIRC. Inductors saturate, so are awesome in low level positions, awful passing greater current. I haven't even touched on noise, of which there are many types.

I found that Polycarbs are very good as inter-stage coupling , I use a 10uF/6V3 from WIMA the devices at high working voltage i.e. like the Philips type at 63V are not so good at low signal levels. I also bi-pass them with 100nF polystyrene type at 50V working.
But for DC decoupling and high frequency harmonics on PSU's I love the ITW Pakron 20uF Capstick type.

[Wonfor14]

Yeah, so the fundamental problem is that asking a single active device to amplify voltage doesn't work very well. BJTs have horrible distortion characteristics, FETs less horrible, valves less so again and can be linear over a small region. Nelson Pass's site has a good explanation on these different devices. Anyway, people being the clever things they are create all sorts of circuit combinations to try a mitigate these non-linearities. There is local degeneration, trading voltage gain for linearity. There is feedback which samples the output, inverts it and feeds it back into the input by all sorts of extremely clever arrangements. There are distortion cancelling arrangements where one devices distortion cancels the other. Each of these has drawbacks. Trying to isolate the 'sound' of a single component in all this complexity is fraught. Capacitors for example, good old polycarbonate caps are awesome when used as main voltage bypass, but less so when used as simple interstage coupling caps. Something to do with the dielectric polarity IIRC. Inductors saturate, so are awesome in low level positions, awful passing greater current. I haven't even touched on noise, of which there are many types.

I found that Polycarbs are very good as inter-stage coupling , I use a 10uF/6V3 from WIMA the devices at high working voltage i.e. like the Philips type at 63V are not so good at low signal levels. I also bi-pass them with 100nF polystyrene type at 50V working.
But for DC decoupling and high frequency harmonics on PSU's I love the ITW Pakron 20uF Capstick type.

Noise in all components that contain current movement capabilities , semicons metal. Just a note here resistor with a value of above 10K seem to be worse for noise, so the bigger the conductor more random electrons more noise, but putting components in parallel can and often reduce there noise, i.e. the early multi transistor MC stages, And at lower frequencies the noise can go up due to thermal noise, slow electrons, I had to design a amp for seismic detection it had to work from 10mHz to 10Hz, and the only way I could get good gain was to use a transformer, OK it would have been massive and the winding noise would be shit. So I mpx the signal at 100KHz and got Sowter to make transformer with 1:100 ratio, and then de-mpx after the secondary. OHM the company loved it, Seamap in Somerset made thosands of the beasts.

[Geoff.R.G]

Geoff if you could find away to put pic up here I would love it please.

I'll see what I can do

[r3xj0hn570n]

But for DC decoupling and high frequency harmonics on PSU's I love the ITW Pakron 20uF Capstick type.

Oooo, they look nice.

[r3xj0hn570n]

I had to design a amp for seismic detection it had to work from 10mHz to 10Hz, and the only way I could get good gain was to use a transformer, OK it would have been massive and the winding noise would be shit. So I mpx the signal at 100KHz and got Sowter to make transformer with 1:100 ratio, and then de-mpx after the secondary. OHM the company loved it, Seamap in Somerset made thosands of the beasts.

Frequency Division Multiplexing? Sounds a lot like digital sampling at 100kHz, Colin. :^Q

[Wonfor14]

Frequency Division Multiplexing? Sounds a lot like digital sampling at 100kHz, Colin. :^Q
[/quote]

No more like a MC1310 stereo decoder, I used a 4016 in a ring modulator in the front end. OK it is digital a bit but after the decoder it was analogue.

[Wonfor14]

But for DC decoupling and high frequency harmonics on PSU's I love the ITW Pakron 20uF Capstick type.

Oooo, they look nice.

Did you see the massive SMD ones with 10 legs mmm lovely and the ripple current ideal maybe in Crossover for speakers, what do you think?? And maybe by using Arnold Magnetics Square Law cores you could build in a very good protection circuit.
I have used them a few time's in SMPSU (at Advance Electronics) and in a 10KW Class D, no semi's to go to silicon heaven, cool.