Agree, transconductance amps might "equalize" full range drivers because of their inherent LF and HF impedance peaks.
However, Elsinore is a 2,5 way speaker with a flat impedance curve; there is nothing to "equalize".
Wether or not current drive has benefits driving flat impedance loudspeakers has been the subject on other threads before.
Defenders and opponents will never agree
.
An interesting discussion and I invite you to read two short articles (also illustrated) by Mr. John Lenard Burnett:
Amplifiers: Solid State amps verses Valve amps
Amplified music : Solid-state valve amp
Amplifiers: Solid State amps verses Valve amps
Amplified music : Solid-state valve amp
The main benefit is distortion reduction in the speaker. The voice coil is in the feedback loop and if you study the subject deeper you will understand that some distortion is "visible" in the current in a voltage driven speaker.
If you like me, is using an accelerometer for further reduction of distortion, you also get frequency compensation. So using CFB and MFB in combination is like 1+1=3
Recommended reading: https://rmsacoustics.nl/papers/whitepaperMFBdesign.pdf
If you like me, is using an accelerometer for further reduction of distortion, you also get frequency compensation. So using CFB and MFB in combination is like 1+1=3
Recommended reading: https://rmsacoustics.nl/papers/whitepaperMFBdesign.pdf
Yes. It all comes down to the interaction of the speaker impedance and the amplifier output impedance.
A FR is a 1st choice for a current amp because of there generally non-reactive impedance. The rise near the fundemental resonance means more power is output so the FR rises where the impedance rises. But it is a tricky balancing act. That is why you see the use of filters in the Pass article so as to make the impedance rise, Rout interaction flattish wrt frequency response.
One can do much the sam ething if you have a variable Rout amplifier where you can dial in just the right aount of Rout to create a balanced output. With drivers with a rising impedance at HF one can also extend the top end. Some speakers like the Fostex FExx6 series are expected to be used with higher Rout amplifiers for this very same “EQ”.
Speakers with passive XOs very rarely pay attention to the impedance/reactive load that they make for the amplifier. There are exceptions. The Elsinore Mk6 and a 3rd party XO for some Klipsch (Crites?) create multiway speakers with quite flat impedance so they care little about Rout.
A current amp can also be applied in an multi-amplified multiway since one is (usually) driving the lousdpeaker driver over a limited range where its impedance is flat(tish). The bass driver still presents an issue, there are few woofers with a low enuff Qm to have an impedance suitable for current drive unless some other measure is taken to deal with a too-much increase in the bass (ie the nature of the enclosure or EQ of some sort).
dave
The elephant in the room isn't getting addressed. The impulses seem to be showing response differences. Is the current amp being used for its EQ?
As I mentioned before, and it was rejected, I do compare two amps after I eq each in the listening position to perfectly flat fr response in the listening position. I do this every time I make change, no matter how small.
So impedance changing fr response argument is not valid.
I am just comparing two amps, all thing being equal. Current drive wins. Gives me chills how good the harmonic profile is, like the clarinet or oboe is right there in the air.
Albinoni, or Telemann oboe concertos are unbelievable. Or restored Bach. I need to relisten all my thousands of cd's and lp's.
Voltage amp comes flat and boring.
So those who claim distortion differences, please provide data.
No, in my case current amp is not doing any eq, I use dsp for that.
As I explained.
How can we move on with so much misinformation. One guy says its voltage amp, other says its zobel. Look: the purpose of feedback is to correct the signal on the output to match it to the signal on the input. Voltage amp corrects it, but it only corrects what's on the output, and it does not care what happens to the signal on the speaker, its like a blind painter. At least current amp is doing something totally different, reading what is happening on the speaker itself. Isn't that important? I am no expert, but with 50+ amplifiers at my place, I can tell you, these sound most like music.
So impedance changing fr response argument is not valid.
I am just comparing two amps, all thing being equal. Current drive wins. Gives me chills how good the harmonic profile is, like the clarinet or oboe is right there in the air.
Albinoni, or Telemann oboe concertos are unbelievable. Or restored Bach. I need to relisten all my thousands of cd's and lp's.
Voltage amp comes flat and boring.
So those who claim distortion differences, please provide data.
No, in my case current amp is not doing any eq, I use dsp for that.
As I explained.
How can we move on with so much misinformation. One guy says its voltage amp, other says its zobel. Look: the purpose of feedback is to correct the signal on the output to match it to the signal on the input. Voltage amp corrects it, but it only corrects what's on the output, and it does not care what happens to the signal on the speaker, its like a blind painter. At least current amp is doing something totally different, reading what is happening on the speaker itself. Isn't that important? I am no expert, but with 50+ amplifiers at my place, I can tell you, these sound most like music.
Last edited:
Adason, I don't disagree.. in fact I'd like to see more reason to agree.
Secondly what is the end result difference between using response EQ to compensate for the amp reacting to the varying impedance, and conjugating the impedance so there is absolutely no need to EQ?
This should give a perfect comparison but why not measure at the amp outputs with a response sweep? Listening rooms are such finicky things.
Secondly what is the end result difference between using response EQ to compensate for the amp reacting to the varying impedance, and conjugating the impedance so there is absolutely no need to EQ?
post #489
most of the time people like the sound of current amp and attribute it to more power in lower frequencies or highs due to increasing resistance getting more power
but i have few systems where I use 31 band digital eq to equalize flat fr response in my listening position
I do this for every change in the system, so comparing apples to apples, I replace application note amp with this transconductance amp, and the result is day and night
Last edited:
Interesting...more distortion with current amp.
I measured 0.049% thd on one amp, and 0.050% on the other, so i considered that no difference. However, i noticed intermodulation distortion was double the value in current amp. Not sure why.
Wonder if simple single sine wave is representative enough of complex music.
I am not sure i follow. I noticed different eq needed for different amp, but that does not surprise me.
However, i noticed big improvement in sound quality with current amp. Not due to fr response differences. I do not measure much difference on sine wave thd.
Problem is, i do not have any way to evaluate it except listening to music. Music is way more complex than one sine wave. I am not trying to sell anything here, just saying we are far from proper way of evaluating amplifiers.
It seems to me that phase is playing role here.
Remember long time ago we had crt color tv? There were three electron beams which hit different light fluorophores. If they were not aligned properly, image sucked. But once it was in sync, it resulted in nice image.
Only if those two impulses are fully identical you will have apples vs. apples. Per definition phases will be identical under that condition. Sorry, but there is no misinformation here as far as I can see.
The impulses can be FFT 'd in order to obtain data in the frequency domain.
Different impulse=different .frd, whether one likes that or not. In case your equalization yields a different impulse, you still have a differences in the frequency domain or phase differences.
Furthermore, a basic Zobel will not flatten the impedance around speaker resonant frequency, but only above some 200Hz.
BTW: It might help the discussion to show your speaker's frequency responses.
The impulses can be FFT 'd in order to obtain data in the frequency domain.
Different impulse=different .frd, whether one likes that or not. In case your equalization yields a different impulse, you still have a differences in the frequency domain or phase differences.
Furthermore, a basic Zobel will not flatten the impedance around speaker resonant frequency, but only above some 200Hz.
BTW: It might help the discussion to show your speaker's frequency responses.
I see what you mean. I measured thd of electric signal on small fullrange using RightMark Audio software.
I can measure distortion using umic and rew, I have that on other laptop with soundcard. Not sure when I get to do it though.
Hello everybody .
in theory perhaps (we must then see it in practice) there is an alternative way to drive the speakers in current, I am not an expert so please correct me if I'm wrong.
in order of appearance:
DAC output stage with current amplifier instead of current / voltage converter. immediately after there is the passive crossover filter in series configuration driven in current and lastly the final amplifier, also in current, in place of the transconductance final. no crossover filter between power amp and speakers.
in theory perhaps (we must then see it in practice) there is an alternative way to drive the speakers in current, I am not an expert so please correct me if I'm wrong.
in order of appearance:
DAC output stage with current amplifier instead of current / voltage converter. immediately after there is the passive crossover filter in series configuration driven in current and lastly the final amplifier, also in current, in place of the transconductance final. no crossover filter between power amp and speakers.