Bob Cordell's Power amplifier book

snapshot

Stability of CFA PAmp...

Here's a web link for any one in need of a very tiny, easy and basic windows snapshot program, it's a very old but still probably one of the best despite its age.
Download WinSnap 1.1.9 - FileHippo.com
v 1.1.9 is the last and completely unrestricted version until the developer changed policies, has worked for me for many years on Win XP.
Thankfully Filehippo saves all the old versions. :)

Another open source alternative can be found here.
Greenshot
 
Disabled Account
Joined 2012
Waly ----->

According to a newer TI: https://www.ieee.li/pdf/viewgraphs/current_feedback_vs_voltage_feedback_amplifiers.pdf

"The fast 'modern' VFB Amp or VFA uses the same architecture as the CFA (in order to reduce the power consumption while maintaining a large SR)". Therefore, it is a CFA topology with buffered (-) port.

The feedback element is constrained in its Z range since it is the compensation element.

The CFA Power Amp curves shown have a flat and very low distortion vs power and vs freq. A quality which IMNSHO lends a Ultra High Def quality to the sound. So, what margin of SR is needed to achieve this flat very low distortion vs freq vs power in a 200W/8 or 400W/4 PA?

THx- RNMarsh
 
Last edited:
"The fast 'modern' VFB Amp or VFA uses the same architecture as the CFA (in order to reduce the power consumption while maintaining a large SR)". Therefore, it is a CFA topology with buffered (-) port.

There is nothing modern about this. Even the Otala amp used heavy degeneration which increased the linear range at the input and improved the slew rate to GBW ratio. Killing the OLG was only a side effect but of course no amount of rational argument convinces the worshipers. Use of degeneration to modify the behavior of op-amps is old, old, old.

BTW ALL conventional op-amps are "current mode" the input stage delivers current in response to an input voltage, you know transconductance.

You can find dumbed down fuzzy thinking and even nonsense on many data sheets. My favorite from BB, FET's are good as op-amp inputs because they are square law and produce more pleasing even harmonics. Wrong, an LTP made with any devices is balanced and its output current is an odd function with respect to input voltage hence produces only odd harmonics.
 
Last edited:
Disabled Account
Joined 2012
....your just making a nice mess of it all, Scott.

I'm getting out and walking... .


Yungas-Road.jpg


Be back in the 1st world in a few weeks - if i make it.


THx-RNMarsh
 

Attachments

  • Leh-Manali-Highway.jpg
    Leh-Manali-Highway.jpg
    108.1 KB · Views: 248
Last edited:
The CFA Power Amp curves shown have a flat and very low distortion vs power and vs freq. A quality which IMNSHO lends a Ultra High Def quality to the sound. So, what margin of SR is needed to achieve this flat very low distortion vs freq vs power in a 200W/8 or 400W/4 PA?

I'm sorry, but at this point any coherent discussion is impossible. You need first to understand the slew rate concept and its correlation (or lack, thereof) with other amplifier parameters. The book Mark recommended is a good start, but it is not at the "management" level where you enjoy placing yourself. Some effort beyond purchasing the book is required.
 
Sorry for the delay Bob, my computer died again. I suspect my PSU keeps destroying motherboards... D:

My amplifier is a VFA with about 22MHz ULGF. The prototype is on the bench and has been running for months without a single strange pop or buzz. I've driven it to clipping and it handles full power in short bursts, the heatsink connection is not really up to extended power yet.

I never said local oscillation is dependent on the amplifier feedback loop. I never dismissed the possibility of legitimate local oscillation. You may have to reread my post. If we are not clear on what exactly the other did or didn't say, then we can't really have a meaningful conversation. I don't use "may not" to deny, and I don't use "probably is" to assert truth; merely to convey probabilities. I try to be very deliberate about how I write.

It's true that cascaded EFs can cause local oscillation, but in astx's amplifier it's a single EF driving Lfets, the liability is much less.
 
Disabled Account
Joined 2012
I'm sorry, but at this point any coherent discussion is impossible. You need first to understand the slew rate concept and its correlation (or lack, thereof) with other amplifier parameters. The book Mark recommended is a good start, but it is not at the "management" level where you enjoy placing yourself. Some effort beyond purchasing the book is required.

I understand it just fine. is there a point you want to make about SR that i might not understand. Let us all know your point.


THx-RNmarsh
 
I understand what Waly is saying, in his own characteristic way. Slew rate margin doesn't always have a direct relationship with distortion. Just like some amps have soft clipping and others have hard clipping. Slew rate is also an overdrive behavior so perhaps you could say "soft slewing" and "hard slewing" is applicable here. So it's not as simple as deciding on a universal margin and expecting it to work on any given design.

To clarify my previous post, there are many situations where the amplifier feedback loop does result in what appears to be local oscillation, which will go away if you disable the feedback loop. I am not saying this is true in all cases, but it is more common than you would think.
 
Disabled Account
Joined 2012
Billshurv --- I believe that is true re Mega slew rate. I never said there was. he keeps saying I think so. Say that enough and it becomes true??

However, I will answer the question he wont/cant. The SR Margin. It is 2X minimum. Or, if you 'believe' that 1v/v output is a good number for determining SR, I am saying it is twice that amount. How much would that be Ref to the CFA power amp shown?

If you can get that characteristic of ultra low distortion and flat across freq and power and load using ANY topology you like to play in, and the SR has a 2X margin, then you have the beginnings of a great audio amplifier.



THx-RNMarsh
 
Last edited:
AX tech editor
Joined 2002
Paid Member
However, I will answer the question he wont/cant. The SR Margin. It is 2X minimum. Or, if you 'believe' that 1v/v output is a good number for determining SR, I am saying it is twice that amount. How much would that be Ref to the CFA power amp shown?

THx-RNMarsh

Richard are you implying it is NOT 1V/V (or twice) for a CFA?

Jan
 
Disabled Account
Joined 2012
Besides the characteristics of low, flat distortion vs freq vs power vs Z, and good SR margin .... I would add best isolation from output to Vas. Or double, triple or v/Lfet ops. Then add in PSRR or effective supply immunity and rejection of emi/rfi from supply or else where, no funnies on over load and the like.... eventually we get close to perfection in audio PA.

The Dadod amp met all my criteria at reasonably high power with reliability and is rock stable.

For me, i just happen to be used to CFA to get there.... starting back in the 70's and never used anything else since. but to each his own... what ever you are comfortable with to get you there. But for a book it would be nice to also show example of CFA or 'modern VFA' (LOL).


THx-RNMarsh
 
Last edited: