Current feedback - Voltage feedback, how do I see the difference?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Disabled Account
Joined 2012
Are the formula's for gain the same? They should be if a CFA is a VFA.

They are sufficiently different to describe a fundamentaly different mode of gain operation.

Thx-RNMarsh

VA-CM formulas.jpg
 
Last edited:
Are the formula's for gain the same? They should be if a CFA is a VFA.

They are sufficiently different to describe a fundamentaly different mode of gain operation.

In general no, because of the loading effect, affecting the open loop gain, which in turn may affect (if not large enough) the closed loop gain. However, in certain conditions (both external - gain and internal - topology), the CFA closed loop voltage gain expression can be the same as for a VFA. Look at the LM6172, a CFA in disguise. Of course, then the main characteristic of a CFA, that is the gain independent bandwidth, is lost. You can't have cake and eat it too.
 
Of course, then the main characteristic of a CFA, that is the gain independent bandwidth, is lost.


This so-called "gain indepedent bandwidth" is an electronic sleight of hand because for low closed loop gains a so-called "CFA" sacrifices forward path gain, and, therefore, loop gain to obtain it.

So, attempting to increase negative feedback (loop gain) by reducing the amplifier's closed loop gain accomplishes virtually nothing in terms of overall linearity because the amplifier's first stage gain is simultaneously reduced.

So, of what utility are so-called "CFAs" in audio applications?
 
Are the formula's for gain the same? They should be if a CFA is a VFA.

They are sufficiently different to describe a fundamentaly different mode of gain operation.

Thx-RNMarsh

View attachment 338158


For a first order analysis, the values of the feedback components appear in the equation for the forward path gain in so-called "CFA", whereas this is not the case in a VFA.

However, this doesn't change the fact that so-called "CFAs" are really compromised VFAs.

Read the section entitled "over compensation" in the figure you posted.
 
Last edited:
Did you actually look into that instead of just following your unreliable gut? Do you realise that the current into the low impedance node is NOT determined by its low impedance?

I did.

Focus Mr. Didden, I would appreciate if you refrain in putting words in my mouth. Did I mention anywhere that "the current in the low impedance node is determined by it's low impedance"? Or, as a matter of fact, did I mention any "low impedance" at all?

All I said is that a CFA is an amplifier with VFB, where the FB network load at the input node cannot be ignored. As Michael says, this affects the open loop gain, which in turn affects the open loop crossover frequency. Lowering the crossover frequency, combined with higher reverse feedback transfer leads to bandwidth independent closed loop gain.

If Michael doesn't like CFA's for audio (for still unknown reasons), this is his own problem. If Mr. Didden really wants (for still unknown reasons) to consider CFA a marketing created brand, that's also his own problem. I'm trying to show that this entire debate is about semantics only.
 
Disabled Account
Joined 2012
For a first order analysis, the values of the feedback components appear in the equation for the forward path gain in so-called "CFA", whereas this is not the case in a VFA.

However, this doesn't change the fact that so-called "CFAs" are really compromised VFAs.

Read the section entitled "over compensation" in the figure you posted.

I have the whole book... do you?

I developed a Current Feedback topology before it was an IC and published it. I got under .001% with it at all audio freqs and it had a bandwidth of >20Mhz at 22v p-p -- that was several decades ago. In fact, the designers at Elantec called me as they were moving in a similar direction and had filed a patent and were pretty upset when i published a similar design/topology.

Since then the current-mode feedback as it has become named has been refined. They are super good, now. But they leave the vfb behind as you move higher and higher up in frequency. They are now a staple in all analog IC mfr'ers portfolio as their strengths overlap with vfb designs. Its above or below the over-lapping areas (grey) where one shines brighter than another.

Its important to know when to use one or the other and why.

Thx-RNMarsh
 
So, attempting to increase negative feedback (loop gain) by reducing the amplifier's closed loop gain accomplishes virtually nothing in terms of overall linearity because the amplifier's first stage gain is simultaneously reduced.
Not necessarily.

In the text Richard uploaded, the discussion of equation 4 is perhaps a bit misleading. While the second factor is independent of gain, it is dependent on RF, the feedback resistor. The statement "This shows that the bandwidth becomes independent of the set gain G" assumes that G is set by changing RG.

If instead, we choose to alter G by changing RF, then this no longer holds true, and the graphs of open loop and closed loop gain vs frequency become the same as for a VFB amp.

In a nutshell, when G>>1 then, to a first approximation:
  • Open loop gain is roughly proportional to 1/RG
  • Closed loop gain is roughly proportional to RF/RG
Thus they can clearly be set independently of each other.
 

Attachments

  • cfb-txt.GIF
    cfb-txt.GIF
    25.2 KB · Views: 118
Disabled Account
Joined 2012
perhaps you havent been using the best model for current-mode feedback?

Need better models.jpg

1. lower noise potential because the devices are in parallel with each other (push-pull);
2. More consistent sound quality character because (?) the distortion's harmonic structure is the same at all frequencies -> at low-mid-high.
3. Very low distortion with only moderate levels of gnfb.
4. Very fast-high speed and wide bandwidth.

THx-RNMarsh
 
Last edited:
What reasons would you advance for using so-called "CFAs" at audio frequencies?

Michael, you obviously haven't read my post here: http://www.diyaudio.com/forums/soli...urrent-feedback-amplifier-10.html#post3416244

I'll save you a click:

I would concede there's no good reason to claim that CFAs are anywhere superior to VFAs for audio, but to state that CFAs are inferior that's, to me, a biased stretch.

Are you sure you are not trolling? 'Cause I've had enough of this.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.