I have not looked at this in the context of the original article. The MJ481/491 output transistors used had unusually high current gain and these have been obsolete for a long time - probably before the advent of SPICE modelling.
Anyway there is an alternative link to the original articles http://www.keith-snook.info/wireles...orld-1970/15-20W Class AB Audio Amplifier.pdf
At the top of page 323 the feedback factor is stated to be 46dB typical - frequency range unspecified.
At the bottom of the same page Fig.7 the "Influence of signal frequency on distortion (1W into 15 Ohms) shows an increase by a factor of 10 between 1kHz and just over 20 kHz - suggesting a decline in negative feedback factor of 20 dB.
MJL3281 and MJL1302 were the best modern power transistors for current gain but considerably less than MJ481/491 so replacing these would demand more base current from the preceding stages and reduce the load presented to the voltage amplifier stage.
Modern speakers have lower impedance than 15 Ohms so the case to increase drive currents is compelling.
However the gain of a common emitter amplifier is a function of Gm roughly 40 per m.A. of collector current multiplied by the collector load in k.ohms. Which adds weight to the idea of increasing current in the Vas and driver transistor.
I have done that in the simulation.
The other change I made is to reduce the compensation capacitor. The 30MHz MJL's will allow more feedback with less phase shift at high frequencies. The value of 270 pF was chosen as I have these in stock - 330 p.F. is more common than 390 p.F. so these values can be used in the simulation if you wish.
When did the bottom half of quasi-comp output become called a "Sziklai"? I suppose the name it's no less useful than "Darlington". I've seen it called a "White buffer" and you could call it a minimal CFA (current feedback amplifier). My point is that these handles are not sacred, so use them if you like but it is what it is and insisting on you version of correctness betrays anecdotal knowledge rather than real understanding.
I suppose it must be since about the 1950's, when Sziklai patented it. It's certainly not a new term.
Naming things is important. If we don't all agree on a name for something, then trying to talk about it will lead only to misunderstanding.
It's a matter of scholarship, as to whether patent terms and names become widely used and for how long. Doubtless, many folk had difficulty spelling or pronouncing the guy's name properly so they avoided it, asserting their own arbitrary choice of terms instead. That's the cue for confusion and why it's better to stick to precedents and singular, unique terms.
An alternative term that is used here and has descriptive value, is CFP or Complementary Feedback Pair. "compound pair" is another old term but it carries no distinctive meaning in itself. Sziklai Pair | Compound / Complementary | Radio-Electronics.com
An alternative term that is used here and has descriptive value, is CFP or Complementary Feedback Pair. "compound pair" is another old term but it carries no distinctive meaning in itself. Sziklai Pair | Compound / Complementary | Radio-Electronics.com
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.