Expecting this right your description.
The geometry of a sound is exactly composed of these parameters as "attak, sustain, decay etc. .." to proceed with our argument, we omit the Dolby system.
It is sufficient to modify the geometry of a sound or sound envelope (if more complex) to move away from the fidelity of an instrument, in the range of 0.5% -1% in the proportion of values which constitute (after interpolation), the shape of the geometry.
Can an amplifier or DAC change this geometry?
Yes, we have to consider that the geometry (in its form) is dictated by the proportions of level between one point and another. (not by the absolute value of amplitude).
A DAC can distort up to 3% this geometry, and this depends on the clock frequency, clock stability, interpolation filter with low bw and a very bad sync if it is connected to a CD player or as a source eg. external CPU.
Amplifier:
If it is a good class AB and has a low hysteresis between the unbalance (when still, modulating) and balance (unfortunately this parameter is not measured) or it can produce a change in the timbre of the instrument. but normally we think the difference is due to the frequency response of the amplifier, if a snare seems less dry.
Last edited:
finally, recent studies (even with my participation) have shown that the PSU is responsible for the modification of the structure of a sound or envelope right in the dynamics relating to musical instruments with fast attack. this has a relationship with the SPL produced also. today it is perfectly demonstrable.
A reason to see 2x150K-uF in a big class AB with seriously transformer. you want to try to replace with 2x15K-uF and listen?
A reason to see 2x150K-uF in a big class AB with seriously transformer. you want to try to replace with 2x15K-uF and listen?
You might want to try a capacitance multiplier circuit rather than humonguous and expensive 150,000 uF caps. Or use a nice SMPS where the noise is in the 400 kHz which will get filtered by the LC filter on the class D amp anyway. If one is going to go class D in amps, there is no point in sticking with big transformer and big caps conventional power supply. The best power supply would be lead acid batteries disconnected from charging unit when playing music. Not sure why more folks don't do this? It has huge high current headroom (400 amps) for deep deep bass. I used to use batteries strictly for ultra low noise phase-sensitive pre-amps for detecting nanovolt signals buried in 100 mV of background noise.
If I had a dollar to spend to improve sound quality, it will go to the speakers , rather than a tube amp. All todays amplifiaers are good, and differnces are marginal. Speakers still can be improved...
+1 on that. The sound quality is probably 75% speaker design and 25% amplifier. Speaker design, driver designs, and simulation software are all still improving. We are so much better than we were 10 years ago. It is similar to the revolution that occurred with optics and lenses after computer ray trace simulations of camera lenses were available. Prior to that, the best lenses came from high end lens manufacturers with a reputation and a price to go with it (Leica being prime example). Following good optical simulation programs, much less expensive lenses are available from most major manufacturers that exceed the performance of those lenses of yesteryear before computer sims. Now with HornResp, MJK, AkaBak, WinISD, etc. the performance of speaker designs can pretty much be predicted without cutting wood. Class-D amps is starting a paradigm shift on what we should expect out of modestly priced and compact amps and will have performance comparable to very expensive amps of yesteryear.
I also believe loudspeakers are the most important link in the chain.
I imagine that over the last 60+ years of Hifi that every conceivable speaker box shape, internal volume and contruction has been produced already, without any software help.
Speakers have changed very little over the years that is why a few select vintage units demand such very high prices on the used market.
I loved my old Tannoys ever more after hooking them up to a very inexpensive d-amp in what was going to be just and experiment.
I'm a d-amp addict, I cannot stop buying em.
I imagine that over the last 60+ years of Hifi that every conceivable speaker box shape, internal volume and contruction has been produced already, without any software help.
Speakers have changed very little over the years that is why a few select vintage units demand such very high prices on the used market.
I loved my old Tannoys ever more after hooking them up to a very inexpensive d-amp in what was going to be just and experiment.
I'm a d-amp addict, I cannot stop buying em.
Hi,
Just to conclude, on your statements:
capacitance multiplier: it is very far from what we have written, becouse is incapable of transient response. (or not know)
Nice smps: I do not know a single one that I can put in a big class AB and has a fast response to transients.
I agree on the batteries then!
regards
That's been discussed a lot here in Class-D, but probably before you joined.
Batteries have low impedance at DC, but that can rise rapidly with frequency. Depends a lot on the battery, of course. Good regulated supplies, either linear or SMPS often work better than batteries.
Batteries are great for some noise sensitive circuits, not so great for power amps.
For fast rise times the current is supplied by any local decoupling caps, rather than the average current, which is supplied by the PSU proper. This is somewhat redundant for standard analogue audio though (ignoring class D) as most transients are going to be incredibly slow compared to any proper switching events. Providing the PSU is capable of delivering the required current, be it linear or SMPS, it doesn't matter. If it cannot supply the required current the rails will collapse, the PSU might shut down and the amp will clip. If this isn't happening then the PSU is delivering enough current to accurately convey any musical transient.
We Men, like the theory because it allows us to speak and speak ...
But in this case it is very easy to see the practical fact.
A toroidal 1kW at 80V with 2x47.000uF has a voltage drop of up to 15-18V repeated under a simple 33-66ms burst @ 1Khz, in a MOSFET amplifier.
below 100Hz, the drop in voltage reaches 20V. thinks with vocals and bass.
I understand that this topic is preferable go around rather than accept it.
But in this case it is very easy to see the practical fact.
A toroidal 1kW at 80V with 2x47.000uF has a voltage drop of up to 15-18V repeated under a simple 33-66ms burst @ 1Khz, in a MOSFET amplifier.
below 100Hz, the drop in voltage reaches 20V. thinks with vocals and bass.
I understand that this topic is preferable go around rather than accept it.
AP2,
Perhaps someone else with more knowledge of the math can chime in here but that voltage drop sounds extremely high for a regulated power supply. At the same time you are not saying if this is a 20 watt amplifier or a 150 watt amp and that is going to make a world of difference to the voltage drop. Were is Gootee when we need him?
Perhaps someone else with more knowledge of the math can chime in here but that voltage drop sounds extremely high for a regulated power supply. At the same time you are not saying if this is a 20 watt amplifier or a 150 watt amp and that is going to make a world of difference to the voltage drop. Were is Gootee when we need him?
Not quite the story. The cilia are not frequency specific but the location of their (hair-)cells on the basilar membrane is relatively frequency-specific and that endows their signals with a frequency significance. In addition, that significance is only relative as many hair cells are excited to varying degrees even by a single pure tone because the on basilar membrane is dispaced maximally but not exclusively as the travelling wave from a pure tone passes from the base of the cochlea up towards the apex.
I am glad that someone intervenes, the problem in the forums is always the same. many are full of theory or slogan of the company and do not know the reality. this is my deduction in relation to the various threads and replies I read.
I followed the thread with interest Gootee (I also posted the curves of a ybride module that can perfectly solve the problem of power supply, unfortunately has a high cost of production).
We started from the recognition of a sound to reach the conclusion (in my case). yes if voltage not decrease, changes everything listening especially in D Class.
my previous post, it refers (+ /-80V) to get in 450w mosfet amplifier AB with very soft attak of instrument, instead of original wonderful dry as original.
Just for curiosity,my post on Gootee thread: http://www.diyaudio.com/forums/power-supplies/216409-power-supply-resevoir-size-108.html
regards
I followed the thread with interest Gootee (I also posted the curves of a ybride module that can perfectly solve the problem of power supply, unfortunately has a high cost of production).
We started from the recognition of a sound to reach the conclusion (in my case). yes if voltage not decrease, changes everything listening especially in D Class.
my previous post, it refers (+ /-80V) to get in 450w mosfet amplifier AB with very soft attak of instrument, instead of original wonderful dry as original.
Just for curiosity,my post on Gootee thread: http://www.diyaudio.com/forums/power-supplies/216409-power-supply-resevoir-size-108.html
regards
Last edited:
Just for curiosity,
This is the fast power regulator (30Amp) (one per side need) that produced the behavior shown in the measurement.
This power module (in my opinion excellent), has not been produced.
there are 5 samples only for testing, has developed on ceramic wafer.
Very simple to use eheh!
regards
This is the fast power regulator (30Amp) (one per side need) that produced the behavior shown in the measurement.
This power module (in my opinion excellent), has not been produced.
there are 5 samples only for testing, has developed on ceramic wafer.
Very simple to use eheh!
regards
Attachments
Last edited:
I am apologizing for not answering right away.
I really like Graaf gear (Graaf.it and in USA Atmosphere is following similar approach). For example the upper BW of 13.5II is overkilling 500KHz. Yes human ear cannot hear ultrasound like 500KHz for sure but this is the most natural sounding preamp I've ever owned. True magic. I know that ultrasonic ICs like OPA627 637 are also have good reputation among audiophiles.
While believing in Fourier I am not treating 20-20.000 Hz approach as practical one (wavelets vs Fourier are another can of worms). Class A tends to deliver better BW even in case when printed specs for an amp are limited to 20-20.000Hz. This my explanation why some audiophiles are leaning towards A - the better BW.
Hi Efficiency could be achieved not only by horns but using line arrays as well. McIntosh are making such speakers for example.
Personally I was owning full size Magnepans and Krell FPB monoblock 600 watts of Class A total. Maggies are the best speakers I've heard so far (guy who had owned Maggies but upgraded later to MBL claims that MBL even better but they both finicky in setup and to amplify). Yes back loaded horns are not as accurate. But I've sacrificed Magnepns in favor of Class A amplification.
Yes HiEnd gear is expensive and rare. Speakers price tags & WAF are limited factors in most cases.
Hi, If you want to listen to Class D with PRAT and base to die for listen to the Texas Ind range of :
TPA3130D2 15 watts
TPA3118D2 30 watts
TPA3116D2 50 watts
I am building 3116 and 3118 as we speak...
Have you built your own pcb or using eval board?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.