Hello everyone, and if you are watching, hello Nelson Pass!
My introduction into audio was auditing a class by the late Marshall Leach Jr. at Georgia Tech. Dr. Leach was famous for his 1976 paper, "Build a Low TIM Amplifier." He set up a blue print and standard that was used for decades.
http://leachlegacy.ece.gatech.edu/papers/lowtim/feb76feb77articles.pdf
Nelson Pass' research and products stand in some ways as polar opposites to this approach however. Where Dr. Leach proposed a recipe consisting of a number of necessary stages to building a successful solid state amplifier, Pass has gone the other way, and attempted to reduce the stages as much as possible.
Can we use this paper and Pass' design philosophy to compare and contrast the state of the art today? As importantly, what changes in manufacturing, especially transistors, has improved that enables Pass-like designs to work without the pre-Leach distortion issues of the first SS designs?
Best,
E
My introduction into audio was auditing a class by the late Marshall Leach Jr. at Georgia Tech. Dr. Leach was famous for his 1976 paper, "Build a Low TIM Amplifier." He set up a blue print and standard that was used for decades.
http://leachlegacy.ece.gatech.edu/papers/lowtim/feb76feb77articles.pdf
Nelson Pass' research and products stand in some ways as polar opposites to this approach however. Where Dr. Leach proposed a recipe consisting of a number of necessary stages to building a successful solid state amplifier, Pass has gone the other way, and attempted to reduce the stages as much as possible.
Can we use this paper and Pass' design philosophy to compare and contrast the state of the art today? As importantly, what changes in manufacturing, especially transistors, has improved that enables Pass-like designs to work without the pre-Leach distortion issues of the first SS designs?
Best,
E
Last edited:
Most of your questions can be answer by fact that Leach amps are class AB and most NP amps are class A.
I think there are a minimum number of stages you can have in a good amplifier. Below that it doesn't function as well as it could do.
On the other hand I have seen some terribly bloated designs with dozens of transistors. I am not convinced they give better results than a good correct design.
There is a happy medium where the amp functions well but doesn't take up too much real estate. There comes a point of diminishing returns.
On the other hand I have seen some terribly bloated designs with dozens of transistors. I am not convinced they give better results than a good correct design.
There is a happy medium where the amp functions well but doesn't take up too much real estate. There comes a point of diminishing returns.
Leach amp design - specially the Super Amp kicks one serious butt though and the parts used justify because of M.Leach Jr. design objectives. Practically has been the basis of most BJT amp design in it's class, until today.
However, NP design has been appealing for DIY hobbyist due to it's simplicity which encourage everyone to try themselves new idea and NP's generous efforts to keep the DIY spirit alive and flaming in this forum.
However, NP design has been appealing for DIY hobbyist due to it's simplicity which encourage everyone to try themselves new idea and NP's generous efforts to keep the DIY spirit alive and flaming in this forum.
If we are honest Leach has only one very popular and well documented DIY project - Leach amp(low tim) - super amp ( too complicated ? )and others were never so popular.Someone does not like that he does not use CCS or speed up capacitor, others do not like bias diodes or zeners in front end...Still listening to my diy Lechs (the oldest one i built 8 or 9 years ago - it was my first DIY amp)
Pass designs have a variety of options...BJT, Fets, simple circuits,Class A, AB etc advanced circuits etc...lots of articles - but there are those who dislike the Pass concepts and philosophy...
the most popular is F5 imho - simple,symmetrical and good sound.I'm still listening my DIY F5, B1 and AB100.
It's a hobby for me - thanks for all those who are open for DIY community.
I have had personal contact with both designers (email or forum) - their attitude and humanity are also the ones that inspire me.
Pass designs have a variety of options...BJT, Fets, simple circuits,Class A, AB etc advanced circuits etc...lots of articles - but there are those who dislike the Pass concepts and philosophy...
the most popular is F5 imho - simple,symmetrical and good sound.I'm still listening my DIY F5, B1 and AB100.
It's a hobby for me - thanks for all those who are open for DIY community.
I have had personal contact with both designers (email or forum) - their attitude and humanity are also the ones that inspire me.
Yes, there are three different ways for constructing an amp today.
First - KISS, keep it simple stupid.
Second - PDFB, possible deep feedback.
Third - SD, simply digital.
First approach results in a DIY friendly and easy-to-build designs. Some of them are popular due to good sounding, but most of them are forgotten in history.
Second approach aren't DIY friendly and demands for good understanding of the feedback theory. The most sophisticated engineers develops an audio amps with techniques that used mostly in aerospace or HF-radio. The goal there is to have as deep fedback as possible while keeping amp from oscillating at any load and signal conditions. The lath is very high now, we can discuss around something like 100-120 dB depth at 20 kHz.
Third approach from a development perspective aren't DIY at all. Now there are well known chips like IRS2092 or TPA3116d2, but the modern amps closes feedback loop in digital domain via fast ADC's. This approach tends to developing not a standalone amp, but to create and properly tune reproduction system at a given circumstances and premises.
Do not look for offensive words where they do not exist.
😉
When in university i was asked about calculating volume of the frustum there were at least three answer variants.
1. Difference between full cone and truncated part
2. Integral over height from base area to top area
3. Integral around radius of base area to radius of top area under integral over height.
At any discipline there are at least three books: primer or ABC-book, tutorial or schoolbook and thesis or dissertation.
First two only helps to understand, last are done by yourself.
First approach to calculating is from primer, second from schoolbook, third you'll develop when you come to work, but what's next?
Severe male would calculate volume as surface integral over closed surface of the frustum!
😉
When in university i was asked about calculating volume of the frustum there were at least three answer variants.
1. Difference between full cone and truncated part
2. Integral over height from base area to top area
3. Integral around radius of base area to radius of top area under integral over height.
At any discipline there are at least three books: primer or ABC-book, tutorial or schoolbook and thesis or dissertation.
First two only helps to understand, last are done by yourself.
First approach to calculating is from primer, second from schoolbook, third you'll develop when you come to work, but what's next?
Severe male would calculate volume as surface integral over closed surface of the frustum!
A few considerations, Pass knows what Leach did. Leach used to update his semiconductor choices as parts became obsolete and finally even with parts going away Pass has better choices.
Now Leach made major strides in amplifier design theory and topology, anyone clever and intelligent can build on that work. Someone foolish my build differently but not better.
Pass is clever and intelligent.
However I have seen designs from others that try to improve on Leach and just don't get it.
Now Leach made major strides in amplifier design theory and topology, anyone clever and intelligent can build on that work. Someone foolish my build differently but not better.
Pass is clever and intelligent.
However I have seen designs from others that try to improve on Leach and just don't get it.
Last edited:
Interesting articles :
http://leachlegacy.ece.gatech.edu/papers/stepatten/stepatten.pdf
W. Marshall Leach, Jr.
Best ragards 🙂
http://leachlegacy.ece.gatech.edu/papers/stepatten/stepatten.pdf
W. Marshall Leach, Jr.
Best ragards 🙂
The value of the Leach paper was not in 1 amplifier. His proposal was used as the blueprint for thousands of commercial designs from that day forward. From Amber to Yamaha. The paper set a standard and recipe book that was followed note for note. If you know how that amplifier works, you can read almost any schematic to a linear amp with BJTs from that day forward. It was a revolutionary point in time. Having high distortion or high TIM was no longer an acceptable compromise in solid state design, because some one had laid out the path and the design criteria for all to follow.
Also, the idea that the differences between Leach and modern Pass are merely A vs. AB is short sighted. You could easily make a leach amp that was biased class A but still maintain all the topology. In fact, you probably had to until the MOSFETS arrived. I don't think Pass class A amps bear much of a resemblance (but would love to discuss).
My point is surely NOT to say one engineer was better than the other. But rather, that time has marched on. How much better are we, how much is design, how much is parts? The Leach amp had no access to FETs at the time. Some of Nelson Pass' designs depend on them.
Has Pass hit on circuit ideas which would have worked in the early 1980s?
I point these ideas out for fun, not to attack one designer or their work over another, but to talk about how the slow evolution of our hobby.
Best,
E
Also, the idea that the differences between Leach and modern Pass are merely A vs. AB is short sighted. You could easily make a leach amp that was biased class A but still maintain all the topology. In fact, you probably had to until the MOSFETS arrived. I don't think Pass class A amps bear much of a resemblance (but would love to discuss).
My point is surely NOT to say one engineer was better than the other. But rather, that time has marched on. How much better are we, how much is design, how much is parts? The Leach amp had no access to FETs at the time. Some of Nelson Pass' designs depend on them.
Has Pass hit on circuit ideas which would have worked in the early 1980s?
I point these ideas out for fun, not to attack one designer or their work over another, but to talk about how the slow evolution of our hobby.
Best,
E
Last edited:
Professional soldering station , amplifiers with triode semiconductors
and bunch of : line stages without global feedback, buffers,
open baffles, full range drivers, horns , high efficiency , autoformers ,
light dependent resistor volume control, Mr. Pass articles and
Burning Amp Festival talks and amplifier kit's for DIY etc etc.
Best regards 🙂
Attachments
Yes !
Very definitely a "must read" paper.
And re-read regularly until it is committed to memory.
It should be compulsory reading for all Members that want to contribute to the "Amplifiers" sections of this Forum.
Also this ones:
https://core.ac.uk/download/pdf/4872885.pdf
https://kenkundert.com/docs/cd2001-01.pdf
No, just light exam for some member rights like ability for long posts.
I jokingly suggested an exam some years ago.
I don't think the idea went down well.
R.D.Middlebrook could do with looking to Leach for ideas on how to present a paper for the less knowledgeable.
Similarly M.Tian et al is just too complicated as an introductory paper for Newbies, Beginners and early Intermediates.
Both papers are great for Members thinking about studying for EE qualifications.
Similarly M.Tian et al is just too complicated as an introductory paper for Newbies, Beginners and early Intermediates.
Both papers are great for Members thinking about studying for EE qualifications.
Last edited:
Yes, that's common university practice - tell so complicated how you can.
The specialty is an conspiracy of experts against newcomers.
But them both have good and easily understandable examples in LTSpice library.
It's absolutely necessary to differentiate members for those who could estimate feedback nominals (or something enough complex like two-pole compensation) from who couldn't.
- Status
- Not open for further replies.