hypex ncore

[pos]

Nice design!
Would (1) SMPS600 and (2) nc400 fit the enclosure with a simpler IEC connector?

[ChrisPa]

Originally Posted by Esperado

ChrisPA, a class D amp is a width modulation of a square pulse. The width is set by a comparator between a triangle signal and the input signal. Its precision (so its definition or distortion) depend of the internal noise and the level of the input signal (you get better precision if you compare V than µV).

Reducing the voltage of the pulses (reducing the voltage feeding the power Mosfet) reduce the power (less voltage) without changing the widths of the pulses. The output noise will be reduced in the same time.
It is obvious.

It will not clip 'earlier' with the input signal. It will just reduce the acoustic power,

This might be the case if the amp ran open loop - but it doesn't run open loop and so it isn't the case. The feedback loop of a class D amp is arranged so that the output signal tracks the input signal (times the closed loop gain) independently of the power supply voltage. Indeed, up until clipping - ie. when the duty cycle hits 100% or 0% there (nominally) will be no difference in the output signal with variation in PSU voltage; there is only a difference in the maximum power available.

If what you described were the case, then the amp would have a 0dB PSRR. The nCore's PSRR is specified as 75-80dB.
The nCore's is specified PSU voltage range is 35-75V; the gain does not change with the supply voltage.

(it is obvious?)

I think you're making some assumptions about the open loop configuration of all class D amps. It's not up to us to tell the designer how to implement the internal loop of the amp ("The width is set by a comparator between a triangle signal and the input signal"). The only thing that is relevant is that the width is set so that the output signal is compared with and tracks the input signal

Originally Posted by Esperado

To explain the things with other words, instead of using 0 to 5% of the modulation range in normal work, you will use 0 to 30% for the same acoustic level. See what i mean ?

no, not at all

Originally Posted by Esperado

About compression driver, the membrane is damped by the high pressure a little like if it was in water. The little volume of air between the phase plug and the membrane oppose a great force against movements. The excursion of the membrane is very little, for the same acoustic level compared at the one you would need in free air.
http://www.meyersound.com/news/press...yer_diag.l.gif

Maybe I should have asked this bit the other way round. I can't understand how "an acoustically damped device" can be "a high efficiency microphone"

[Esperado]

Sorry, ChrisPa, but i have no courage to expose here the class D theroy of operation. There is a lot of papers about.
Pulse-width modulation - Wikipedia, the free encyclopedia
Global feedback in class D works like all feedbacks, comparing the output signal with the original one. The main difference between the Bruno's work and the others is he included the output filter in the loop, reducing coil distortion that way, not an easy work because of the poles.
This has nothing to do with the manner the pulse modulation is done, and, sorry to tell-it, you seems to do not understand.
As well as the influence of the DC on the power switching devices: a class D amp has 0dB PS rejection if no feedback !!!!!
The level set by each pulse is function of its width AND of the level of the switched DC. Some even adds a little class A analog amplifier at the bottom of the pulses.

About, high efficiency, i wrote that a high efficiency speaker is a high efficiency microphone. DOT.
Dynamic mics or speakers are reversible. I used some loudspeakers in drum kits as microphones for stage PA, and some used mic cells as tweeters.
Please do not fake my inputs.
It is frightening to see how in this forum some like to contradict for pleasure.

[robbbby]

I just read the last 10 or so pages to catch up and I am beyond confused.
Anyways I just purchased some new speakers, revel ultima studio 2. With my old speakers I was just running them biwired from the ncore. Would there be any benefit to picking up two more modules and running my new revel speakers bi amped?

What about biamping with using one smps per 2 ncores versus biamping and using a smps for each ncore?

I know it's been brought up before but we're 600+ pages deep now, and i'm about 6 strong beers deep.

[OllBoll]

Originally Posted by robbbby

I just read the last 10 or so pages to catch up and I am beyond confused.
Anyways I just purchased some new speakers, revel ultima studio 2. With my old speakers I was just running them biwired from the ncore. Would there be any benefit to picking up two more modules and running my new revel speakers bi amped?

What about biamping with using one smps per 2 ncores versus biamping and using a smps for each ncore?

I know it's been brought up before but we're 600+ pages deep now, and i'm about 6 strong beers deep.

The advantage is if you use crossovers before the NCores thus doing away with big passive components on the speaker side. Otherwise it should not make any difference at all, the NCore already has stupidly low distortion so any distortion you get is from the speaker and the speaker _only_. So while in theory you get lower distortion by separating it with 2 NCores per channel in practice it will only hurt your bank account.

[robbbby]

My bank account cannot take any hurt at the moment. It was not something I was really planning on doing, more of a curiosity thing. With such a big upgrade i'm double checking all my other equipment to make sure it's up to par. Thanks for the feedback!

[OllBoll]

Originally Posted by robbbby

My bank account cannot take any hurt at the moment. It was not something I was really planning on doing, more of a curiosity thing. With such a big upgrade i'm double checking all my other equipment to make sure it's up to par. Thanks for the feedback!

You're welcome =)

The real "problem" with passive crossovers is that it is harder to do what you want and if you go low the component costs skyrocket unless you want to compromise even more but it's not bad if it's done well. You can get away with more things by going active and using pre amp crossover but another option is to design the speaker to compensate this limitations which they have probably already done so the improvement should be miniscule if any at all if they did it right.

The main advantage with going active with DSP is that it would be so easy to get it good enough, not that it would be better except for saving time

[Juhleren]

@robbbby,
Bruno has claimed that biwiring a passive speaker all the way back to the ncore terminals should (i guess theoretically) offer what biamping can due to the high DF.

biamping (also passively) should provide improved working conditions for each amp though in the particular case of ncore i don´t think it has been empirically established yet. One claimed that the audible load-dependency between a serial and a parallel connection of his speakers relied more on the cables he used. That could indicate that Bruno´s statement above convey to practice also...

@OllBoll,
Passive and active filters are two very different animals with each their pro´s and cons. Personally I would think the entire construction of a speaker very differently depending on whether I went down the active or the passive road. (remember that the attenuation you get from passive filters depends on the impedance of the connected drivers -which varies according to frequency and coil displacement) -active opens a very different can of snakes...

btw, did you ever fix the tweeter XO on your diy carlsson clones as we talked about?

best,

[OllBoll]

Originally Posted by Juhleren

@OllBoll,
Passive and active filters are two very different animals with each their pro´s and cons. Personally I would think the entire construction of a speaker very differently depending on whether I went down the active or the passive road. (remember that the attenuation you get from passive filters depends on the impedance of the connected drivers -which varies according to frequency and coil displacement) -active opens a very different can of snakes...

btw, did you ever fix the tweeter XO on your diy carlsson clones as we talked about?

best,

That thing was fixed ages ago =)

What remains though is the problem that the tweeters shouldn't really be crossed that low and the main driver can't be crossed much higher and so to fix that you would need 4th order XO which would be beyond a passive design.

In the end it still works good enough at moderate levels but I will ensure that my oncoming speaker I'm designing myself won't have this issue =)

[Juhleren]

Originally Posted by Esperado

The level set by each pulse is function of its width AND of the level of the switched DC. Some even adds a little class A analog amplifier at the bottom of the pulses.

Not to claim any authority on these matters but isn´t the gain exactly "controlled" by the nfb loop...
If you run on lower DC the fb loop just compensates the signal so that you get accordingly wider pulses and thereby the same output level relative to the input -hence the same gain.
From what I understand the distortion goes up for the same output if you lower DC because the pulses have to become wider than if you ran with higher DC... The THD curve moves from left to right when rail DC goes up.

best,