I can report that there is no buzz at all on my CFA's.
You can look at the PSRR both statically and dynamically.
With no input signal you should get zero noise on the speaker output - that's static PSRR.
Drive you amp hard at 25 or 30kHz into a speaker.
You should not hear any LF component in the speaker. That's dynamic PSRR and a much more important indicator of your amps PSRR performance.
Take care BTW that you don't burn your speaker out. You can also just drive a dummy load and hook you speaker up with a dropper resistor.
If you do hear anything, you will need to work out if is common impedance coupling or a real PSRR problem.
You can look at the PSRR both statically and dynamically.
With no input signal you should get zero noise on the speaker output - that's static PSRR.
Drive you amp hard at 25 or 30kHz into a speaker.
You should not hear any LF component in the speaker. That's dynamic PSRR and a much more important indicator of your amps PSRR performance.
Take care BTW that you don't burn your speaker out. You can also just drive a dummy load and hook you speaker up with a dropper resistor.
If you do hear anything, you will need to work out if is common impedance coupling or a real PSRR problem.
Last edited:
A bandwidth extending to 200KHz will introduce no more than 6 degrees of phase shift at 20KHz; this is insignificant.
What is all the fuzz about? You derive the feedback signal from a voltage divider in both #1 and #3. The current source in #1 can only be used to set up the bias current since it is in parallel with a much smaller resistance.
There is no magic in sacrificing open loop gain for wider band width. All the circuits I have seen so far in this thread is about using a single-ended or long-tailed-par input. The LTP is achieves lower THD by cancelling some effects of Collector to Emitter current on the Base to Emitter voltage. It comes at a price, lower band width.
Which sounds the best? I haven't a clue, why should there be a difference?
To me it somewhat amusing the read in postings that we should ignore THD since the THD of speaker elements are magnitudes bigger but at the same time we should care about slew-rate, what is the slew-rate of tweeters?
There is no magic in sacrificing open loop gain for wider band width. All the circuits I have seen so far in this thread is about using a single-ended or long-tailed-par input. The LTP is achieves lower THD by cancelling some effects of Collector to Emitter current on the Base to Emitter voltage. It comes at a price, lower band width.
Which sounds the best? I haven't a clue, why should there be a difference?
To me it somewhat amusing the read in postings that we should ignore THD since the THD of speaker elements are magnitudes bigger but at the same time we should care about slew-rate, what is the slew-rate of tweeters?
Sorry to repeat but I think my question was overlooked:
Is the supposed CFA's higher distortions at low frequency because of non-linearities outside of the input stage (i.e. caused by output stage distortions within the CFA lower loop gain) ?
That is, in isolation does the CFA input stage (compared to an LTP with all else equivelent and ideal) still have higher low freqeuncy distortions.
Thanks
-Antonio
Is the supposed CFA's higher distortions at low frequency because of non-linearities outside of the input stage (i.e. caused by output stage distortions within the CFA lower loop gain) ?
That is, in isolation does the CFA input stage (compared to an LTP with all else equivelent and ideal) still have higher low freqeuncy distortions.
Thanks
-Antonio
"CFA" can be more linear at large diff Vin - in the extreme they can operate Class AB where long tailled pair would saturate
at very low level diff input V, corresponding to high loop gain, the input stage differences are more in implementation detail - not fundamental topology - everything becomes more linear with less signal
but the practical advantage of matching like type semi's may favor diff pair over complementary inputs
at very low level diff input V, corresponding to high loop gain, the input stage differences are more in implementation detail - not fundamental topology - everything becomes more linear with less signal
but the practical advantage of matching like type semi's may favor diff pair over complementary inputs
Last edited:
I believe it would become a 50Hz buzz.... Of course the voltage regulator would prevent it all the same
At least for the stereo module I built, I can crank out 30KHz into the speaker and no LF artifacts can be heard. This proves that the CFA topology referred to here can be made to have excellent power supply rejection, static and dynamic.
Agree - regulator can make an important difference. Both my VFA e-Amp (active ripple eaters in the rail) and my CFA sx and nx-Amps use heavy filtering in the rails. For the ripple eaters, you can kill both the LF and HF gunk. The simple RC filters get rid of the HF stuff - there's scope pics of both in the articles on my website.
JCX,
Thanks I understand what you are saying, for purposes of this thread I was trying to dig deeper into differences between CFA and VFA.
The same question phrased somewhat differently, would the low frequency distortion of a CFA change (input stage distortion only) if the feedback impedance was lowered (same ratio and again ideal components) ?
Indeed the CFA would have much less distortion as it enters AB operation compared to an LTP clipping. At least for audio purposes I would think both of these extremes would be avoided and hence the inapplicability of the CFA slew rate characteristic.
Thanks
-Antonio
Last edited:
Interesting how distortion (and noise for that matter) must blindly be as low as possible, yet slew rate and bandwidth only need to be 'enough'
Fascinating, isn't it?
Deaf guys rely on THD screen to be sure the amp is OK, others listen and are bothered if PRAT is not OK. Fascinating, isn't it?Interesting how distortion (and noise for that matter) must blindly be as low as possible, yet slew rate and bandwidth only need to be 'enough'
Just why would you want more bandwidth or slew than you need?
Mikeks,
Same can be said for ultra low distortion also....
Deaf guys rely on THD screen to be sure the amp is OK, others listen and are bothered if PRAT is not OK. Fascinating, isn't it?
THD specs on paper are more fascinating for some people....
Yes , they are symmetrical , but not enough for my taste....
I m not anti CFA but rather against extraordinary claims.
VFA can also be considered as CFA but with a follower buffer
CFA is more direct feedback, just one transistor to realize the comparison of error signal. VFA is actually using a buffered copy of feedback to compare it with original signal. But VFA is more symetric, because the original and the feedback path is the same, even it is not required for amplification.
There are many circuit realizations, some even with both inputs buffered with follower(like LM3886). I like simple realization whereever posible. Simpler circuit mostly will have the wider bandwidth, if properly designed, have no worse linearity. Best of all, it can be more throughly understood. IC is usually more complicated to suit for cirtain specifications, they mostly have lots of sacifices made to make them safer to use, thus their performance are not at their best, otherwise their should be even better.
CFA is more direct feedback, just one transistor to realize the comparison of error signal. VFA is actually using a buffered copy of feedback to compare it with original signal. But VFA is more symetric, because the original and the feedback path is the same, even it is not required for amplification.
There are many circuit realizations, some even with both inputs buffered with follower(like LM3886). I like simple realization whereever posible. Simpler circuit mostly will have the wider bandwidth, if properly designed, have no worse linearity. Best of all, it can be more throughly understood. IC is usually more complicated to suit for cirtain specifications, they mostly have lots of sacifices made to make them safer to use, thus their performance are not at their best, otherwise their should be even better.
Think it's important to present here the info from the other thread where "blameless" fans are gathered, blindfolded to the other's opinions. Again my statement follows: current feedback amplifiers outruns VFA in all parameters important for audio signals amplifying. This statement is based on years of experiences, testing and listening ready made amps from known producers as well as from all kind of DIY amps made from both topologies. Attached three steps describing what VFB actually is - to some "upgraded" current feedback circuit, while logical thinking will lead you to use wire only (fig1.) instead of nonlinear part (fig2.) if you pursuit an audio quality in power amplifiers. Case closed.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.