I had read Nelson's paper previously and there was one thing that confuses me with Nelson's paper and another with Mark's post. First is that Nelson's last sentence of the paragraph is: "If you invert the phase of the signal before the circuit, then it will keep the positive 2nd." (context based on the first sentence in the paragraph?)
I have to think that it was intended like this: If you invert the phase of the signal before the circuit, and you reverse the polarity at the speakers, then it will keep the positive 2nd. Is that correct?
Then Mark with his confusing #1 and #4 results in his post that I don't see a way to achieve:
"1. Negative phase H2, end-to-end signal path is inverting."
"4. Positive phase H2, end-to-end signal path is noninverting."
Below is what I determine is the result of Mark's examples:
(input jack) --> (gain = +1) --> (H2 gen) --> output jack = Audio phase inversion and positive H2
(input jack) --> (gain = -1) --> (H2 gen) --> output jack = Audio phase not inverted and negative H2
(input jack) --> (H2 gen) --> (gain = +1) --> output jack = Audio phase inversion and positive H2
(input jack) --> (H2 gen) --> (gain = -1) --> output jack = Audio phase not inverted and negative H2
(input jack) --> (gain = +1) --> (H2 gen) --> (gain = +1) --> output jack = Audio phase inversion and positive H2
(input jack) --> (gain = +1) --> (H2 gen) --> (gain = -1) --> output jack = Audio phase not inverted and negative H2
(input jack) --> (gain = -1) --> (H2 gen) --> (gain = +1) --> output jack = Audio phase not inverted and negative H2
(input jack) --> (gain = -1) --> (H2 gen) --> (gain = -1) --> output jack = Audio phase inversion and positive H2
Please correct me if I am wrong or have missed something.
I have to think that it was intended like this: If you invert the phase of the signal before the circuit, and you reverse the polarity at the speakers, then it will keep the positive 2nd. Is that correct?
Then Mark with his confusing #1 and #4 results in his post that I don't see a way to achieve:
"1. Negative phase H2, end-to-end signal path is inverting."
"4. Positive phase H2, end-to-end signal path is noninverting."
Below is what I determine is the result of Mark's examples:
(input jack) --> (gain = +1) --> (H2 gen) --> output jack = Audio phase inversion and positive H2
(input jack) --> (gain = -1) --> (H2 gen) --> output jack = Audio phase not inverted and negative H2
(input jack) --> (H2 gen) --> (gain = +1) --> output jack = Audio phase inversion and positive H2
(input jack) --> (H2 gen) --> (gain = -1) --> output jack = Audio phase not inverted and negative H2
(input jack) --> (gain = +1) --> (H2 gen) --> (gain = +1) --> output jack = Audio phase inversion and positive H2
(input jack) --> (gain = +1) --> (H2 gen) --> (gain = -1) --> output jack = Audio phase not inverted and negative H2
(input jack) --> (gain = -1) --> (H2 gen) --> (gain = +1) --> output jack = Audio phase not inverted and negative H2
(input jack) --> (gain = -1) --> (H2 gen) --> (gain = -1) --> output jack = Audio phase inversion and positive H2
Please correct me if I am wrong or have missed something.
If this is true, why do you reverse polarity of the speaker on the ACA or the ACA mini (jumper off)? The original ACA squishes the top of the waveform, see post 274 in the ACA thread.
On most speakers, a positive Voltage on the "red" terminal of a speaker moves the cone towards the listener with a few exceptions, see posts 3667 and 3668 on the B1 with Korg Triode thread.
I think it would be awesome if someone could confirm, but below is how I think it would be.
Leave out the speakers (for now). Let's stop at the output jack and (for now) assume that speakers are connected "normally". i.e. when the voltage at the positive speaker terminal is at a higher potential than the negative speaker terminal, then the cone moves outward.
I also need to assume that for Mark's buffers it leaves everything about the incoming signal ideal to output with the exception that it amplifies. i.e. it leaves the relative distortion intact.
So, the H2 generator (normally) generates an inverted signal with a positive 2nd order harmonic. Only when the speakers are hooked up "backwards" does it go to having the output fundamental in phase with the input fundamental. However, the part to lose sleep over is that the relative phase of the 2nd harmonic to the fundamental also "flips".
Then comes the additional fun part... "If you invert the phase of the signal before the circuit, then it will keep the positive 2nd."
So, it's important to note (like Mark did) what's happening at the output jacks vs. what's happening at the speaker and how they're hooked up along with the order.
Let's take one or two examples.
Example one from above.
(input jack) --> (gain = +1) --> Phase intact no distortion
(H2 gen) --> output jack => Absolute phase inversion and positive H2
Example 2 from above.
(input jack) (gain = -1) --> Absolute phase inversion no distortion.
(H2 gen) --> output jack = Audio phase not inverted and positive H2
So, as long as someone else chimes in that I read it properly and interpreted it properly, then I think you can work through the rest. I'm curious if I've had it correct too, but the above is the way I had thought to be correct.
Hope it helps us both.
Leave out the speakers (for now). Let's stop at the output jack and (for now) assume that speakers are connected "normally". i.e. when the voltage at the positive speaker terminal is at a higher potential than the negative speaker terminal, then the cone moves outward.
I also need to assume that for Mark's buffers it leaves everything about the incoming signal ideal to output with the exception that it amplifies. i.e. it leaves the relative distortion intact.
So, the H2 generator (normally) generates an inverted signal with a positive 2nd order harmonic. Only when the speakers are hooked up "backwards" does it go to having the output fundamental in phase with the input fundamental. However, the part to lose sleep over is that the relative phase of the 2nd harmonic to the fundamental also "flips".
Then comes the additional fun part... "If you invert the phase of the signal before the circuit, then it will keep the positive 2nd."
So, it's important to note (like Mark did) what's happening at the output jacks vs. what's happening at the speaker and how they're hooked up along with the order.
Let's take one or two examples.
Example one from above.
(input jack) --> (gain = +1) --> Phase intact no distortion
(H2 gen) --> output jack => Absolute phase inversion and positive H2
Example 2 from above.
(input jack) (gain = -1) --> Absolute phase inversion no distortion.
(H2 gen) --> output jack = Audio phase not inverted and positive H2
So, as long as someone else chimes in that I read it properly and interpreted it properly, then I think you can work through the rest. I'm curious if I've had it correct too, but the above is the way I had thought to be correct.
Hope it helps us both.
Looking at the perception side of things… Perhaps there is distortion cancellation when taking the speaker into account? Consider a very small sealed speaker that displaces 10cc (peak to peak) of air with a 150cc back volume. Assuming perfectly symmetrical displacement and a large room so the front volume is considered infinite: the air compliance changes as the back volume becomes 145cc during the inward stroke and 155cc outward. This would produce second harmonic (dominant) distortion.
This is a pretty well known effect for many speaker design engineers, so often a driver’s suspension compliance is designed to be asymmetrical in free-air then becomes more symmetrical (but not perfect) when placed in an appropriate enclosure. Super symmetry
Also note that this effect varies with amplitude and frequency being most apparent around/below resonance.
Or maybe it’s the opposite and some people just like a little bit of distortion
Also... what's neat... or not (if you're looking for some particular mix of 2nd) is that if one piece of gear in the chain produces positive 2nd and another produces negative 2nd, they cancel. Two box super-symmetry. Joking of course.
That's why I play with the H2 generator mostly with amps and preamps that have really low distortion or known distortion. As ZM would say... no added sugar. It's fun to measure the distortion in the entire end-to-end chain along with each piece and see how they "stack". Still not an expert at it by any means, but Nelson didn't call out any serious errors in my post, so I'll assume I'm on the right track as I plug along.
That's why I play with the H2 generator mostly with amps and preamps that have really low distortion or known distortion. As ZM would say... no added sugar. It's fun to measure the distortion in the entire end-to-end chain along with each piece and see how they "stack". Still not an expert at it by any means, but Nelson didn't call out any serious errors in my post, so I'll assume I'm on the right track as I plug along.
H2-generator
Hello out there,
some parts stuffed. Still thinking about putting the trimpots on the pcb or to
use externally, wire connected trimpots at the frontpanel....
And a little PSU is still necesssary...CRCRC-filter + regulator?
Nice project!
Cheers
Dirk
Hello out there,
some parts stuffed. Still thinking about putting the trimpots on the pcb or to
use externally, wire connected trimpots at the frontpanel....
And a little PSU is still necesssary...CRCRC-filter + regulator?
Nice project!
Cheers
Dirk
Attachments
cubicincher: see my post #84 in the "H2 V2 Build" forum. Using M.J.'s PO89ZB plus a wall wort works wonderfully with Nelson's little circuit.
Cheers,
Pete
Cheers,
Pete
Hello out there,
some parts stuffed. Still thinking about putting the trimpots on the pcb or to
use externally, wire connected trimpots at the frontpanel....
And a little PSU is still necesssary...CRCRC-filter + regulator?
Nice project!
Cheers
Dirk
Hi Dirk,
I have the first version of the H2. I know you like very nice overkills, but TBH as Pete just said: SMPS+ Mark's filter are more than fine for this one.
The orginal H2 is a nice little soundbox that I mainly use as rough indicator at friend's to see if they could benefit from more negative H2. IMHO it does the job fine on normal HIFI systems. It doesn't have though the sonic qualities of a tweaked B1K to enhence more performing HIFI systems being used as a preamp for example. Hence not worth IMHO going for the ultimate PS, the H2 signature and limits will dominate.
That applies to the original H2, no clue on the later version, but seems close enough on the paper...
All IMHO of course
Claude
I have the first version of the H2. I know you like very nice overkills, but TBH as Pete just said: SMPS+ Mark's filter are more than fine for this one.
The orginal H2 is a nice little soundbox that I mainly use as rough indicator at friend's to see if they could benefit from more negative H2. IMHO it does the job fine on normal HIFI systems. It doesn't have though the sonic qualities of a tweaked B1K to enhence more performing HIFI systems being used as a preamp for example. Hence not worth IMHO going for the ultimate PS, the H2 signature and limits will dominate.
That applies to the original H2, no clue on the later version, but seems close enough on the paper...
All IMHO of course
Claude
I put 2 of these boards in series and set both set for unity gain.
The first one is adjusted for no distortion and inverts the signal.
The second is adjusted to 1% as per instructions and inverts the
signal back to the original phase.
I did it that way so the phase would be correct and
easy to switch in and out of the system as desired.
The diferance in the sound is subtle but noticeable,
I would describe it as adding a bit of fullness.
The first one is adjusted for no distortion and inverts the signal.
The second is adjusted to 1% as per instructions and inverts the
signal back to the original phase.
I did it that way so the phase would be correct and
easy to switch in and out of the system as desired.
The diferance in the sound is subtle but noticeable,
I would describe it as adding a bit of fullness.
Last edited:
I did not measure the voltage. I looked at the spectrum display from the program
Spectrum Analysis and adjusted it until there was no 2nd harmonic displayed.
Looking more at this one thing to consider if you change from one power supply
to another, you may need to readjust the pots if the voltage changes between them.
Thinking about this more it may be interesting to wire up switch to
flip the positions of the stages around (clean then 2H / 2H then clean).
Spectrum Analysis and adjusted it until there was no 2nd harmonic displayed.
Looking more at this one thing to consider if you change from one power supply
to another, you may need to readjust the pots if the voltage changes between them.
Thinking about this more it may be interesting to wire up switch to
flip the positions of the stages around (clean then 2H / 2H then clean).