As I see it the Hawkfords error correction is a kind of feedback. But since it measures the error across the output stage, inverts it and adds it to the input it still has some advantages compared to regular NFB. It´s correct that NFB divides the distorsion, but the Hawkfords EC cancels it if the measured error is added in exactly the correct amount. Thats the hard part, since this amount is probably frequency dependant, temperature dependant and so on.
Still, getting rid of 90% of the output distorsion or so, makes it much easier for the global NFB divide the remaining distorsion.
The other advantages like improved PSRR is also nice.
Still, getting rid of 90% of the output distorsion or so, makes it much easier for the global NFB divide the remaining distorsion.
The other advantages like improved PSRR is also nice.
Rikard Nilsson said:
Hi Rikard. Do you mean like the circuit below?
Do you mean like the circuit below?
Yes, in the sense that the global loop needs less feedback loop gain to get the distortion to the target. Of course, the total loop gain of the amp needs to be the same, if you will. So the decision is where to apportion the feedback.
Is this not a consequence of floating the psu of the correction circuit, tying it to the output signal, rather than the type of feedback?
Brian
Attachments
The circuit you attached above is what I mean. Cancelling errors. Very smart, very fast. And I also mean that only because an amp might have EC in the output stage, there is no reason to lower the global NFB simply because there is less distorsion coming from the final stage. For my own amp I plan to keep using a fairly ha NFB as well.
Rikard Nilsson said:The circuit you attached above is what I mean. Cancelling errors. Very smart, very fast. And I also mean that only because an amp might have EC in the output stage, there is no reason to lower the global NFB simply because there is less distorsion coming from the final stage. For my own amp I plan to keep using a fairly ha NFB as well.
Ok, I understand. The circuit I show is entirely a negative feedback circuit and is identical in operation to HEC except that it needs no tuning. I shall coin this "Brian's Error Correction".
Of course, I believe it is commonly referred to as a servo.
I agree that one might as well have as much global feedback as is healthy, all things being equal.
Brian
I prefer not to use the term "cancel" with any feedback circuit as the actual behaviour is asymptotic reduction of error. A feed forward circuit, on the other hand, really does do cancellation in a mathematical sense.
Shahriar said:
Which "Classic Amplifier" are you referring to?
The use of supply rails as current feeds goes back several decades. I think I first came across it in a Meridian amplifier in the 1980s.
As a means to generate two, opposite phase, output currents using a single component (IC), it is clever. It is convenient to be able to set the transconductance by way of the output impedance to ground.
Some things to be wary of in this topology are the phase imbalance between the rails and the non-linearity introduced by all of the op-amp circuit, and the added complexity of a local psu for the op-amp. Note that even if we pretend the output voltage of the op-amp is distortionless, the internal currents that are required to counter the transistor distortions will appear as psu currents.
Good convenience, but I would avoid this, as a general rule, for very high performance circuits. Jan's use of the AD844 is probably a pretty good choice as it appears to be quite symmetrical (we don't have access to the complete circuit of the amp).
Hi everybody
See how much these two amplifier look like each other
http://home.no.net/andiha/articles/audio/cascamp.htm
http://www.ultraaudio.com/features/2006_10_01.htm
Look at the layouts;
I think Vitus has used this amp in bridge mode!
and used some better Components. maybe some better compensation;
What do U think?
Shahriar
See how much these two amplifier look like each other
http://home.no.net/andiha/articles/audio/cascamp.htm
http://www.ultraaudio.com/features/2006_10_01.htm
Look at the layouts;
I think Vitus has used this amp in bridge mode!
and used some better Components. maybe some better compensation;
What do U think?
Shahriar
$5865 for speaker cables!
Anyone selling cables at these prices is not to be trusted, and will write their advertising blurb with no regard for truth.
The only consolation is that anyone fool and self-indulgent enough to spend $50,000 dollars on an amplifier deserves all they get.
It's long past time the advertising standards authorities clamped down on stuff like this.
w
A fool and his money are soon parted.
Anyone selling cables at these prices is not to be trusted, and will write their advertising blurb with no regard for truth.
The only consolation is that anyone fool and self-indulgent enough to spend $50,000 dollars on an amplifier deserves all they get.
It's long past time the advertising standards authorities clamped down on stuff like this.
w
A fool and his money are soon parted.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.