
Home  Forums  Rules  Articles  The diyAudio Store  Gallery  Blogs  Register  Donations  FAQ  Calendar  Search  Today's Posts  Mark Forums Read  Search 
Solid State Talk all about solid state amplification. 

Please consider donating to help us continue to serve you.
Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving 

Thread Tools  Search this Thread 
2nd July 2013, 03:24 AM  #3681 
diyAudio Member
Join Date: Dec 2009
Location: Newnan GA

Kgrlee if Peter Baxandall had used tubes (values) to make his point than really Jurassic devices would have applied . However the use of transistors move at least in to the tertiary devices. Old men have a different view of time than the less experienced folk. There is still time but do pass the beer.

2nd July 2013, 10:05 AM  #3682  
diyAudio Member
Join Date: Sep 2006

Quote:
Cheers, Bob 

2nd July 2013, 10:21 AM  #3683 
diyAudio Member
Join Date: Aug 2009

@ Bob Cordell
Yeah you're right, & i've been saying it for years. Life ain't Linear, it's Logarithmic 
2nd July 2013, 10:46 AM  #3684 
Banned
Join Date: Oct 2010
Location: Chigwell, Essex

There is no such thing as an "all pass" characteristic in respect of the TIS. Period.

2nd July 2013, 12:23 PM  #3685  
diyAudio Member
Join Date: Jan 2011
Location: London

Quote:
I've already mentioned before, the fact that the bipolar transistor current gain Beta exhibits a RHP zero at a frequency larger than the transition frequency Ft is a textbook fact (pp. 6). However, this has no bearing with anything that resembles a TIS in an audio amplifier. The problem is not about abusing the "excess phase" concept, but the underlying attempt of selling the illusion that audio amplifiers have an extra degree of freedom in this respect. In fact, for any minimum phase system like an audio amplifier, once that the "maximum feedback" (in the Bode and/or Lurie sense) has been reached, it is impossible to further increase the loop gain without compromising the stability.
__________________
I am nobody. Last edited by Waly; 2nd July 2013 at 12:25 PM. 

2nd July 2013, 01:22 PM  #3686  
diyAudio Member

Quote:
Clearly nonminimum phase, a surprise to me. It is above the unity crossover frequency but sufficiently close to be of interest. Pink trace is the result of an 820 pF capacitor in parallel with the VAS emitter resistor. I have not yet worked out the details. Circuit is from Toni's amp. 2stageEF high performance class AB power amp / 200W8R / 400W4R Best wishes David Last edited by Dave Zan; 2nd July 2013 at 01:50 PM. 

2nd July 2013, 01:59 PM  #3687 
diyAudio Member
Join Date: Dec 2009
Location: Newnan GA


2nd July 2013, 02:04 PM  #3688  
diyAudio Member
Join Date: Sep 2006

Quote:
I should have used the term nonminimum phase and or with a righthalf plane zero. The term allpass generally implies a flat frequency response, which the VAS obviously does not have. The Millercompensated VAS most certainly does have an RHP zero, albeit at a high frequency. If you feed a Millercompensated VAS with an AC current source, and you assume that the transistor has high beta and high ft, then almost all of the input current flows through Cdom. At most frequencies, the VAS then acts as an inverting integrator, with the collector current of the VAS transistor essentially equaling that through Cdom. To make this collector current flow, a small amount of voltage from the input current must appear at the base, and that voltage is inversely proportional to gm of the VAS. At very high frequencies Cdom begins to look like such a small impedance that the voltage at the base combined with gm does not produce enough collector current to equal that of the current through Cdm. In essence, at high frequencies where Cdom is like a short, the transistor just ends up looking like a resistor whose value is 1/gm. In this case, the applied input current merely appears across this resistance noninverted. The functionality of the Miller integrator has thus transitioned from inverting to noninverting, and is a nonminimum phase device. Millercompensated VASs that are degenerated have lower transconductance, so the frequency at which the impedance of Cdom equals that of 1/gm is lower. With Cdom = 100pF and 1/gm = 25 ohms, that frequency is approximately 60MHz. Cheers, Bob 

2nd July 2013, 02:09 PM  #3689 
diyAudio Member
Join Date: Dec 2006

For people with not enough background I can give hereunder some references explaining better my post; if they do not understand technically they can improve at least their technobabble skills
But first some terminology. An OTA is exactly what Davada has explained. A differential amplifier current mirror loaded is a simple OTA, an amplifier with high impedance input, high impedance output is analyzed as an OTA because high impedance input is a voltage input sensor and high impedance output is a current source so voltage in and current out so transconductance. The first two stage of an opamp is an OTA, a single stage folded cascode is an OTA. They are building blocks in IC design. If the second gm block is Miller compensated it is called a Miller OTA When we analyze stability of feedback system it is custom to call singularities of the transfer function as right half or left half plane located. Root locus is drawn in this plane for a variable parameter usually the DC gain. The plane is the imag, real plan of complex values; drawing impedance or transfer function values for a variable parameter in this plane is called a polar plot in the polar diagram. Roots are called positive or negative in the polar diagram. This is pure terminology.. Stability is now analyzed by considering the positive or negative value of the input impedance or admitance Lundberg (his paper: Internal and external opamp compensation) shows a very nice derivation of a Miller two stage OTA transfer function showing the second order denominator and positive zero numerator. The classical Gray and Mayer are deriving the same equations for a common emitter or a differential amplifier. Roberge ( Operational amplifier : theory and practice ) shows the same and why this positive zero presents a difficulty in the drawing of the root locus and how to solve it. Roberge book is also excellent If you use the drawing and equation of Lundberg for example you will see that the positive zero frequency is fz= gm2/2piC where C is the Miller capacitance and gm2 the second stage transconductance. The dominant pole frequency is f1 =1/KC where K is a constant and the non dominant is f2= gm2/2piCl Where Cl is the loading capacitor This approximation holds for large values of C miller. If you plot the position of the pole/zero with respect to frequency for C as a variable, you see immediately by examination of the given expressions that fz decreases in frequency if C increases while f2 stays constant. At a certain value of C, the zero will be at a lower frequency than the p2 non dominant pole; If you plot now the bode diagram of this transfer function you see first an horizontal from DC then a decrease after the first pole and then again an horizontal after the zero. The phase will go from zero to 90° after the first pole but then instead of going back to zero as with a negative zero, it continues towards 180° because of the sign in the numerator ( sfz) In the region beyond the zero and before the second pole, the amplitude of the transfer function is flat while the phase is gradually going to 180°. This is the charactérisric of an all pass ( in this zone). I am not saying that the stage is an all pass everywhere. 
2nd July 2013, 02:14 PM  #3690  
diyAudio Member
Join Date: Sep 2006

Quote:
Nobody has said that the closedloop response of an audio amplifier is nonminimum phase (at least at frequencies not significantly greater than ULGF). Nobody has said that a nonminimum phase characteristic is a good thing for an amplifier or a useful extra degree of freedom. What was referred to is the openloop gain. Specifically, I noted that the Miller integrator can be a nonminimum phase circuit. Rather than making absolutist statements about what you perceive to have read, you would do better to ask for clarification of what has been said, since there is often a better way that could have been said. Oftentimes, what people say for reasonable brevity may have unsaid caveats. It is perfectly reasonable to bring these caveats to people's attention or give the poster an opportunity to clarify by citing those caveats. This is how reasonable technical discussions are carried out, and everybody learns, rather than being confused by a person picking an argument. Cheers, Bob 

Thread Tools  Search this Thread 


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Another realization of Bob Cordell's THD Analyzer  giulianodes  Equipment & Tools  37  9th October 2013 02:48 AM 
best audio amplifier book??  Bouvett  Everything Else  30  13th August 2012 02:43 AM 
Amplifier Design Book  pixie  Everything Else  27  11th June 2010 08:36 PM 
Project 11.1 from Slone "HighPower Amplifier" Book  Karl71  Solid State  46  6th October 2008 03:47 AM 
bookaudio power amplifier by Douglas Self  mikee12345  Solid State  8  16th November 2003 01:16 AM 
New To Site?  Need Help? 