6th provides a lot of coverage and well worth having.
I always liked 2nd edition. It taught me a lot even though I had read most of D.Self in the technical press.
I still reckon that JLH is a must read. It covers a much wider range of topics, almost everything Linear, but not in such great depth. A good complement to D.Self.
Cordell is probably a "must have". Shame about the digital/switching section. Really needs a separate book, after Cordell has learned more on the topics.
I always liked 2nd edition. It taught me a lot even though I had read most of D.Self in the technical press.
I still reckon that JLH is a must read. It covers a much wider range of topics, almost everything Linear, but not in such great depth. A good complement to D.Self.
Cordell is probably a "must have". Shame about the digital/switching section. Really needs a separate book, after Cordell has learned more on the topics.
I am afraid you understand it wrong. The main noise sources in a power amplifier are the input transistors, their emitter degeneration resistors, the current-mirror, and the NFB network. The current source biasing has a negligible effect.
The noise output of a typical Blameless amplifier is below -92 dBu. Is that not quiet enough?
I suppose you are aware that zeners and LEDs are effectively shunt regulators? What reference did you use for your shunt regulator?
Last edited:
First, while I have been studying power amplifier circuits for decades, I have learned more from your book(s) (and Cordells) in the past year or two than in the previous 4 decades.
I appreciated the chapter regarding the push-pull VAS. This was advocated by W. Marshall Leach some years ago and appeared to be the way to go. But, the actual circuit is not quite as it would appear on paper and as you point out, complementary transistors really aren't, and there is one feedback path attempting to correct 2 signal paths.
I am curious about output topology. It appears that the EF3 configuration might be better behaved in some respects yet your designs for Signal Transfer are CFP. The CFP may be easier to compensate with the bias spreader but has a narrower range. So, is the CFP used in the Singal Transfer circuits because you are better able to properly tune to that narrower range whereas the home builder may be better off with EF3?
I appreciated the chapter regarding the push-pull VAS. This was advocated by W. Marshall Leach some years ago and appeared to be the way to go. But, the actual circuit is not quite as it would appear on paper and as you point out, complementary transistors really aren't, and there is one feedback path attempting to correct 2 signal paths.
I am curious about output topology. It appears that the EF3 configuration might be better behaved in some respects yet your designs for Signal Transfer are CFP. The CFP may be easier to compensate with the bias spreader but has a narrower range. So, is the CFP used in the Singal Transfer circuits because you are better able to properly tune to that narrower range whereas the home builder may be better off with EF3?
I am afraid you understand it wrong. The MAIN NOISE SOURCES in a power amplifier are...
Try it and then decide.
I am not that competant to explain further, but I can tell you I am not the only person who has heard it.
For minimum and quick evaluation, you can use a pair of batteries in an appropriate circuit.
Yes, the circuit I used has a zener for reference.
Gajanan Phadte
Try it and then decide.
I am not that competant to explain further, but I can tell you I am not the only person who has heard it.
For minimum and quick evaluation, you can use a pair of batteries in an appropriate circuit.
Yes, the circuit I used has a zener for reference.
Gajanan Phadte
filtering is good - but putting at cap across a Zener's few Ohms dynamic Z is way inefficient - use a RC low pass
audio weighting curves all show low frequency is very much less audible at low level - 100 Hz corner is good for considerable perceptual noise roll off
and of course amp stage noise contributions are textbook knowledge - diff pair balance knocks down tail ccs noise, mirror degen R is needed to the keep that stage from contributing
VAS ccs noise gets divided by beta on the way to becoming input referred noise for comparison - pretty much out of the picture
audio weighting curves all show low frequency is very much less audible at low level - 100 Hz corner is good for considerable perceptual noise roll off
and of course amp stage noise contributions are textbook knowledge - diff pair balance knocks down tail ccs noise, mirror degen R is needed to the keep that stage from contributing
VAS ccs noise gets divided by beta on the way to becoming input referred noise for comparison - pretty much out of the picture
Mr Self,
Personally I would have liked to see more than a half a page on the Symmetrical Current Feedback input stage, it's an old topology but still relatively untouched. It does require little more than hfe matching to acheive low dc offset, which seemed to be a reason you didn't delve further if I recall correctly?.
Colin
Personally I would have liked to see more than a half a page on the Symmetrical Current Feedback input stage, it's an old topology but still relatively untouched. It does require little more than hfe matching to acheive low dc offset, which seemed to be a reason you didn't delve further if I recall correctly?.
Colin
You're going to have to give me a clue what you're talking about here.
p215 maybe?
That is because it has all sorts of drawbacks, as described, a major one being that it needs a push-pull VAS, which as Chapter 8 shows is not a good way to go.
I'd be glad to hear your thoughts on the push-pull VAS issue- it seems to me that Chapter 8 demolishes the whole business. That was not what I was expecting when I started my research; it's a bit disconcerting when the most conventional approach keeps coming out as the best.
Hi,
Great with an update I'm still on 3th edition, my boss has agreed to let me have one
I miss alternatives to interface between VAS and preamp (opamp like topologies) and better rundown of symmetrical vs. single ended designs.
I also miss designs that works well at lower idle currents in output stage and VAS - optimizing idle losses. Again IC based / power opamp solutions may be worth looking at.
Thanks
\\\Jens
Great with an update I'm still on 3th edition, my boss has agreed to let me have one
I miss alternatives to interface between VAS and preamp (opamp like topologies) and better rundown of symmetrical vs. single ended designs.
I also miss designs that works well at lower idle currents in output stage and VAS - optimizing idle losses. Again IC based / power opamp solutions may be worth looking at.
Thanks
\\\Jens
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.