>They don't have to know that the speackers are reversed. You connect the red terminal to gnd and the black-one to the amp's out inside the housing of the amp. On the outside everithing looks right (not that anything is wrong anyway).
I thought about that, but his speakers have a wire that runs between the 2 speakers. May still work as you say, but I'm unsure what that wire does and wouldn't want to screw something up............mike
I thought about that, but his speakers have a wire that runs between the 2 speakers. May still work as you say, but I'm unsure what that wire does and wouldn't want to screw something up............mike
Hi Guys,
Has anyone tried NIGC with NF resistors of 10k/220k rather than the most popular 1k/20k. If yes did you find any differences in the sound. I do expect them to sound different. And the NF cap should be smaller and better type too.
A few nights ago I ripped out the pre amp section of my Yamaha stereo receiver (9 years old but the power amp topology seams good enough). Connected the output of the input selector to the volume ctrl pot (100k) and from there directly to the input of the power amp section. I have a brand new high-end stereo system now.
The music flows freely from the speakers. The placement of the instruments/vocals is so vivid. Now I can point at them. Before everything was located left, right and center, no in between. Now I can’t tell where the speakers are when I close my eyes. The soundstage has dept and height now. I’m ecstatic about it.
The experiment continued with adding an external classic opamp buffer (OPA2604, gain=1) between the volume ctrl pot (100k) and the Power amp. Guess what. The instruments got grouped more around the speakers and the center and the dept of soundstage changed too (for worse in my case). The overall tonal balance didn’t change much (at least I couldn’t hear difference). And Yes different OPAMPS sound different. I experimented on my CD player.
This experience totally changed my prospective on how the audio systems should be build. I’m absolutely convinced that the less stuff on the way of the signal the better. That’s why the GC seams SOOO appealing to me. And NO active pre amps and stuff like that for me (that’s me though, others may like it better with pre amps, personal taste).
Any way I had to air it out.
Just received my LM3875TFs.
Looking forward to the experience.
Greg
Has anyone tried NIGC with NF resistors of 10k/220k rather than the most popular 1k/20k. If yes did you find any differences in the sound. I do expect them to sound different. And the NF cap should be smaller and better type too.
A few nights ago I ripped out the pre amp section of my Yamaha stereo receiver (9 years old but the power amp topology seams good enough). Connected the output of the input selector to the volume ctrl pot (100k) and from there directly to the input of the power amp section. I have a brand new high-end stereo system now.
The music flows freely from the speakers. The placement of the instruments/vocals is so vivid. Now I can point at them. Before everything was located left, right and center, no in between. Now I can’t tell where the speakers are when I close my eyes. The soundstage has dept and height now. I’m ecstatic about it. The experiment continued with adding an external classic opamp buffer (OPA2604, gain=1) between the volume ctrl pot (100k) and the Power amp. Guess what. The instruments got grouped more around the speakers and the center and the dept of soundstage changed too (for worse in my case). The overall tonal balance didn’t change much (at least I couldn’t hear difference). And Yes different OPAMPS sound different. I experimented on my CD player.
This experience totally changed my prospective on how the audio systems should be build. I’m absolutely convinced that the less stuff on the way of the signal the better. That’s why the GC seams SOOO appealing to me. And NO active pre amps and stuff like that for me (that’s me though, others may like it better with pre amps, personal taste).
Any way I had to air it out.
Just received my LM3875TFs.
Looking forward to the experience.
Greg
I believe some of yall are overestimating the intellegence of the public.I only sell my amps with my horns and instructions have to be as simple as possible.Red out(amp) goes to the red speaker terminal, black out goes to the black terminal.
Even if you crossed them with another set of speakers it would be hard to damage anything, BUT as stated above , i never overestimate the intellegence of the public. Keep it simple.<-my motto.
ron
Even if you crossed them with another set of speakers it would be hard to damage anything, BUT as stated above , i never overestimate the intellegence of the public. Keep it simple.<-my motto.
ron
Most types of opamps have better sonics when used inverted, not just GCs. In the early eighties when the 5534 was all the rage i was quite disgusted by the sound in normal (or should i call it 'milwood') connection. Inverted, it made very decent noises. I eventually built a split riaa phono using 3 inverted 5534s with a disastrous s/n but absolutely great sound.
Konnichiwa,
I did make such tests over 15 Years ago. The inverting connection with the then "new" TDA2030/2040/2050 Series Chip Amplifiers was a biggie. It was NOT a small difference. I experimented afterwards with other Op-Amp's in inverting mode (uA709, NE5532/5534, TL07X/08X Series etc...) finding much the same. Modern high performance Amplifier Chips benefit less from "invert dah suckah" but they still do to a disturbing degree.
Why a "disturbing degree"? Because it is entierly possible to build Amplifier circuits that are free from the issues that cause the difference in the first place. I could think readily of both circuits using strictly N-Channel/NPN Devices and complementary designs, though I'd personally champion the kind of all NPN circlotron circuit I used for my PA Amplifier designs.....
Sayonara
millwood said:
I did make such tests over 15 Years ago. The inverting connection with the then "new" TDA2030/2040/2050 Series Chip Amplifiers was a biggie. It was NOT a small difference. I experimented afterwards with other Op-Amp's in inverting mode (uA709, NE5532/5534, TL07X/08X Series etc...) finding much the same. Modern high performance Amplifier Chips benefit less from "invert dah suckah" but they still do to a disturbing degree.
Why a "disturbing degree"? Because it is entierly possible to build Amplifier circuits that are free from the issues that cause the difference in the first place. I could think readily of both circuits using strictly N-Channel/NPN Devices and complementary designs, though I'd personally champion the kind of all NPN circlotron circuit I used for my PA Amplifier designs.....
Sayonara
Kuei Yang Wang said:
so exactly how did the experiment go? what's the set-up?
Kuei Yang Wang said:
If you believe in Self's experiment on this issue, the difference in THD between non-inverting and inverting is about 0.00003% (or something in that neighborhood).
Maybe we just have different definitions of "disturbing", or Self's data isn't to be trusted.
Peter Daniel said:
I seem to recall that the Great One, Douglas Self, says that there is an audible "phase" tolerance. Does it really matter after a couple glasses of claret?
As for the capacitor on the "+" input, I have left it out and it doesn't seem to matter. I also use a 1uF input cap instead of 2.2uF in the Thorsten design.
I do find it necessary to use a 10ohm resistor with 0.7uH choke on the output.
Jack
Konnichiwa,
Exactly speaking, the experiment went very well.
At the time I worked for a company that manufactured modular mixing desks for TV & Radio. They where still fully discrete back then and I was re-designing some modules. Tests where made with different plug in, routed through the matrix in the board frame, tests where blind (but not ABX protocol).
I was unaware that Mr. Self conducted listening tests, he must have avoided mentioning this. His Measurements are interesting but do not neccesarily have any relevance to the subject of percieved sound.
Neither. Self's data is not relevant to the issue and definition of disturbing is in any good dictionary.
Sayonara
millwood said:
Exactly speaking, the experiment went very well.
millwood said:
At the time I worked for a company that manufactured modular mixing desks for TV & Radio. They where still fully discrete back then and I was re-designing some modules. Tests where made with different plug in, routed through the matrix in the board frame, tests where blind (but not ABX protocol).
millwood said:
I was unaware that Mr. Self conducted listening tests, he must have avoided mentioning this. His Measurements are interesting but do not neccesarily have any relevance to the subject of percieved sound.
millwood said:
Neither. Self's data is not relevant to the issue and definition of disturbing is in any good dictionary.
Sayonara
Kuei,
what about matching the impedances for pos. and neg. input
for the non-inverting configuration? There is an article by
Walt Jung where he seems to claim, although I might have
misunderstood him, that this causes a cancellation of the
common mode induced distorsion. What is you experience/opinion
on this? I know it is not always possible to achieve such a
balance, but when possible, it seems non-inverting config.
would works just as well according to this theory.
what about matching the impedances for pos. and neg. input
for the non-inverting configuration? There is an article by
Walt Jung where he seems to claim, although I might have
misunderstood him, that this causes a cancellation of the
common mode induced distorsion. What is you experience/opinion
on this? I know it is not always possible to achieve such a
balance, but when possible, it seems non-inverting config.
would works just as well according to this theory.
Konnichiwa,
The impedance balanceing does not solve the common mode issue, it adresses a different problem. Bipolar transistors have an appreciable base current which is among other things signal dependent. By balancing the impedances on both inputs you make this the base current modulation in effect common mode.
With the LM3875 the input long tail Pair is fed from emitter followers which will reduce the effect of this by a factor 100 to 1000 (40 to 60db reduction in level), so balancing the impedances is less critical.
Sayonara
Christer said:
The impedance balanceing does not solve the common mode issue, it adresses a different problem. Bipolar transistors have an appreciable base current which is among other things signal dependent. By balancing the impedances on both inputs you make this the base current modulation in effect common mode.
With the LM3875 the input long tail Pair is fed from emitter followers which will reduce the effect of this by a factor 100 to 1000 (40 to 60db reduction in level), so balancing the impedances is less critical.
Sayonara
Kuei Yang Wang said:
OK, I accept the argument for the LM3875. Haven't checked
the datasheet schematic of it lately.
However, unless I am misinterpreting Jung, he does not agree
with your first statement, for op amps in general. Although
balancing impedances matters wrt. input currents, that is not
what he discusses in his article. Rather he specifically discusses
its effect on distorsion caused by the input capacitances. He also
seems to argue that this is even more important for JFET-input
op amps than bipolar ones. I am not clear about whether he
claims such balancing would cancel all common-mode distorsion
or if there are remaining factors still making inverting config.
better.
The article I am referring to is "Walt's Tools and Tips: Op-Amp
audio" from Electronic Design, Dec 14, 1998.
The article seems no longer to be on line, though, but I have
it in case you wish to read it and don't already have it.
Oyasumi-nasai
Christer said:
my wild guess: if you think of a less-than-ideal opamp with junction capacitance as one ideal opamp with capacitance to the ground on both inputs. By balancing out the "input resistance" on both + and - input, you allow the signal on the capacitors remain the same.
Otherwise, since the R/C networks may not charge up and down at exactly the same time, two identical signals on the inputs of a less-than ideal opamp will produce different "signals" on the inputs of the ideal opamp.
Millwood,
Yes, I quite agree there are various imperfections that matter.
Imperfect matching in the op amp is one, although they probably
tend to have very well-matched inputs. Phase shift in the
feedback loop is another, of course.
I guess there are actually several issues to consider here.
1) Assuming a perfect op amp and no phase shift in the feedback
loop, would the impedance balancing cancel all common-mode
distorsion, or are there other factors still causing some
common-mode distorsion?
2) Assuming the answer to 1) is that we would get 100%
cancellation, then how big difference do the imperfections
cause, compared to the inverting configuration. Varies
depending on the circumstances of course.
3) Instead assuming the answer to 1) is that there are also
other factors, then how important are these other factors?
Are they usually very small compared to the capacitance
factor or not?
I really do not how much we are to read into Walt Jungs
claims on cancellation, whether he means we get 100%
cancellation in the perfect case, of it he just means this is
one factor that is cancelled. Walt, if you happen to read this,
your view on this would be much appreciated.
Yes, I quite agree there are various imperfections that matter.
Imperfect matching in the op amp is one, although they probably
tend to have very well-matched inputs. Phase shift in the
feedback loop is another, of course.
I guess there are actually several issues to consider here.
1) Assuming a perfect op amp and no phase shift in the feedback
loop, would the impedance balancing cancel all common-mode
distorsion, or are there other factors still causing some
common-mode distorsion?
2) Assuming the answer to 1) is that we would get 100%
cancellation, then how big difference do the imperfections
cause, compared to the inverting configuration. Varies
depending on the circumstances of course.
3) Instead assuming the answer to 1) is that there are also
other factors, then how important are these other factors?
Are they usually very small compared to the capacitance
factor or not?
I really do not how much we are to read into Walt Jungs
claims on cancellation, whether he means we get 100%
cancellation in the perfect case, of it he just means this is
one factor that is cancelled. Walt, if you happen to read this,
your view on this would be much appreciated.
Christer said:
If you assume a perfect opamp, it already has a 100% rejection of common mode signals, and load-balancing is irrelevant.
the more interesting case is dealing with a less-than-perfect opamp. I imagine the answer critically depends on what you think is causing common-mode distortion.
millwood said:
I think we have a confusion of terminology here, possibly caused
by me.
With perfect op amp, I do not mean the ideal op amp we learn
about in textbooks. I mean a real op amp topology, with real
transistors, thus having non-linear VCE-dependent Cob (or
similar for JFETS), but perfect in the sense that all
components are perfectly matched etc.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.