Single or dual differential?

[PMA]

bocka,

and what to do when there is 2nd harmonic of some -115dB, 3rd of some -120dB, IMD of similar order, no higher harmonics, class A, and you find that it is not THD and IMD what makes a difference, but filtration, CD signal treatment, cable termination etc? That most of the standard approaches to solve sound problems are simply wrong?

[AKSA]

That was a tantalizing post, PMA....


Hugh

[john curl]

Well stated, Bocka. Most of our distortion is in the output, followed by the high voltage driver that has to develop the complete voltage swing for the amp. The input stage used to be a big problem when we used maximum Gm input stages. This was because the input stage would work harder and harder with increasing frequency, ultimately causing slew rate limiting, and even earlier, TIM (or SID). Walt Jung and Matti Otala have published reams of info in this, beginning in the 70's.
Today, we all degenerate our bipolar input stages, increase our gain-bandwidth of our amp, or both. Fets usually don't have to be degenerated in order to get a very high slew rate, because they are always lower Gm than non-degenerated bipolars, and they are more linear as well.

[dimitri]

boska wrote:
Obviously a VAS stage has a high output and typical bipolar stage a very non-linear input impedance witch is the main mechanism of crossover distortion in a bipolar output stage.


Sorry this is true _only_ when the _open_ loop VAS is taken into account. When you speak about Self/blameless design you should have in mind that because of the local loop through frequency compensation capacitor the vas output resistance will _lower_ with frequency. Definitely it wouldn’t be the “main mechanism”.

[fdegrove]

Hi,

What keeps on surprising me is how tube and transistor problems are similar....

Not knowing all that much about semis, you can still read JCs post as if it were tubes he talked about .

It's perfectly understandable to both parties...absolutely brilliant.

Than there's Matti Otala's work regarding TIM and other distortion artefacts that no oscope shows up...

Suddenly I feel at home...great!!!

Cheers,

[jam]

Frank,

Please......you talk like you might build a solid state amp soon.

Here is some fun reading for you....to get in the mood.

http://users.arczip.com/rmcgarra1/xstrhist.html

Regards,
Jam

[fdegrove]

Hey Jam,

Please......you talk like you might build a solid state amp soon.

I do? I wouldn't count on it if I were you...

Let's just say that I'm at awe with JC's knowledge and I really find myself in his thinking and your sense of humour at the same time...

There's still way too much for me in the tube department to be learned, a field that I'm really passionate about, before I even think of looking into semis.

The more I learn, the more I realise I don't know nothing at all...don't even know if that's a proper English expression...

Be my guest and educate this hopeless Rita...

[lumanauw]

I have another "dumb" question. Look at sch1.jpg. I got this design from Randy G Slone book, about the "Optimum Power Amp". The schematic is very nice to see, complementary differential with current mirror, the VAS is cascoded with 2 transistors.
But when I make experiment of it, this design just doesn't work. Up until now I cannot figure what is the voltage in the base of Q11 and Q12 (first VAS transistor). I need it to determine the VAS's standing current. But the base of Q11 and Q12 is located between collector of differential and collector of current mirror. How can I calculate the voltage between 2 collectors?
When I just change the current mirror with ordinary resistor, this amp work. But how come the author have measurement figures, if the amp doesn't work?

Question no2. comes in sch2.jpg. This is the power amp from motorola. Why is the collector pin of Q2 and Q4 doesn't tied up to +/-VCC with it's own resistor? Q1 and Q3 tied up to +/-VCC via 2.21k resistor. In ordinary design, the Q2 and Q4 will go straight to +/-VCC or via another 2.21k. In that design, both collectors are tied to the VAS's resistor, the 22,1ohm resistor. What is the point of this design, and what is the advantages and disadvantages?

[lumanauw]

This is sch2.jpg
*I wonder, yesterday this thread is 5 pages, but this morning is only 3 pages. Why is that?

[Tube_Dude]

Hi Lumanauw!

Up until now I cannot figure what is the voltage in the base of Q11 and Q12 (first VAS transistor).

The voltage in the base of of Q11 is 1,2volts (two diode drops) under the rail positive voltage!
Ex. If the rail is + 40 volts the voltage at the base of Q11 will be 38,8 volts!

In Q12 will be the some but with negatif voltage!(two diod drops from the minus rail)

Question no2. comes in sch2.jpg. This is the power amp from motorola. Why is the collector pin of Q2 and Q4 doesn't tied up to +/-VCC with it's own resistor? Q1 and Q3 tied up to +/-VCC via 2.21k resistor. In ordinary design, the Q2 and Q4 will go straight to +/-VCC or via another 2.21k. In that design, both collectors are tied to the VAS's resistor, the 22,1ohm resistor. What is the point of this design, and what is the advantages and disadvantages?

This is for some inside feedback loop...it can increase stability and linearity..
Advantages..as expressed formely more stability and linearity in open loop...
Disadvantages...less open loop gain for more overall feedback for less overall distortion...
See one example of this technic in a design of our forum friend Jonh Curl:

http://marklev.com/marklev/JC3/jc3schematics.jpg

Regards!