Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › Mini Dac TDA1543 X 4 NOS
New Posts  All Forums:Forum Nav:

Mini Dac TDA1543 X 4 NOS - Page 55

post #811 of 935
Quote:
Originally Posted by Sapientiam View Post

You're tilting at windmills here. I have not so far mentioned anything about roll-off at 22kHz and why would I? That in the range of ultrasonics.

As for the glaring errors on that website - here's one to be going on with. The writer conflates 'multi-bit DAC' with 'R2R DAC'. So then when he goes on to claim that Philips made R2R DACs, he's in error. They didn't. They made some of the best (if not the best) multibit DACs but none were R2R. Want to argue that as a fact? Here's another - he says that the PCM1704 has 'true 24 bit noise-free resolution' - the datasheet disagrees with this claim. So are we to believe BB/TI engineers or the website author?

Wow. Isn't that just a nice insult.

I'm merely giving information. You are not. You are just stating (unfounded) facts and opinions.

The writer conflates 'multi-bit DAC' with 'R2R DAC'.
So what is the difference? What else was there other than R2R before the invention of the sigma-delta DAC? If Philips invented the dac for CD what chips did they make before?

So then when he goes on to claim that Philips made R2R DACs, he's in error. They didn't.
So what are you saying? That the TDA1543 isn't made by Philips or that it is no R2R? Then what is it? Yes, a multibit, but how does it work?

Here is another site; same content (not surprising since they are in the test). enjoythemusic.com
It is clear that mother of tone is not the first with a NOS or ZO-DAC.
I would like to compare my modded Muse 4x with that competition. cool.gif

Burr Brown stateds for its PCM1704: SoundPlus™ 24-Bit, 96kHz BiCMOS Sign-Magnitude Digital-To-Analog Converter.
Here is the pdf where it states 24-bit resolution. and the noise-free you can see in the oscilloscope pictures: it still shows a sinus at -120dB compared to a flat line with a 20bit dac.

This is a quote from a developer (who seems to actually know what he is talking about and is willing to explain in plain English) that says it all imho (src).
Quote:
Again look at the data sheet, and see that the 1704 does 24/96 only (for implied input). But it really does more, however, who had the idea (back then, 1998) to commercialize that ? AFAIK there just was no material higher than 24/96, and so the IDEAS were different because of that reason alone.

The 1704 does work with R2R though (but keep in mind the "sign magnitude" thing which adds a little smartness to it) and for 24 bits there are no others anyway.
Quote:
Originally Posted by Sapientiam View Post

Not to split hairs here but no, reality in general isn't linear. Linear is a special case and almost always an approximation just to make the math easier (like no longer impossible).

Just what reality isn't linear? Man, you are so confusing the issues. How can a violin IRL not sound like a violin IRL? How can reality not be reality 1:1 in normal Newtonian physics?
Yes a loudspeaker has to be linear to make a recorded violin sound lifelike. And so does a headphone. For earplugs it is different however. There the acoustics of our outer ear have to be taken in account and this give a different frequency response. I think the Etymotics website has good information on that.
Edited by ]eep - 3/7/12 at 11:02am

Gear mentioned in this thread:

post #812 of 935
Quote:
Originally Posted by ]eep View Post

I don't mean to argue. I respect you. But it's the way you state things without any proof or links that kind of irritates me.
Actually? No. It's not who you say you are, it's what you do. Convincing is done by reasoning.
You are talking about someone (a German dipl-ing, an education which is held in the highest esteem anywhere in the world) making 'glaring errors' without saying what they are thereby disqualifying someone you don't know anything about who has a lot more to show for himself than you (again, no disrespect towards you). I only have an ing-degree in construction (which includes acoustics btw) so I'm disqualified in this matter anyway. But I do have 30 years experience in hifi and I do know how to listen.
The HF-rolloff you speak of is totally irrelevant. 1dB is the hearing threshold for discerning sound pressure-level differences at 1000Hz , human hearing reaches to 20kHz at a young age, 15kHz at best at my age with hardly any discernment at all as to loudness at that frequency. So 3.9dB difference at 22kHz is moot. Also does no-one state anywhere how much HF-roll-off is at 20kHz (which would be the point where it starts to mean anything) or how much at the extreme of my (or your) hearing.
Look at this:
fadb-f1.gif
See what the dynamic range of a healthy young person is at 20kHz (20dB) compared to 1kHz (130dB)? And I'm not even talking about the range of music (orchestral range in the graph).
fletcher.jpg
This is how 'linear' our hearing is! Again, this is for a young person that hasn't ruined his hearing with his ipod, headphones, concerts etc.
This is how our absolute hearing threshold changes with the years.
The Absolute Threshold of Hearing (ATH) is the volume level at which one can detect a particular sound 50% of the time. If one has a low absolute threshold, it means that he is able to detect small amounts of stimulation, and thus is more sensitive. If one has a high absolute threshold, then he requires more stimulation and thus is less sensitive.
The following are curves of the ATH for several age groups over frequency. Zwicker approximated these curves from experiments on a number of subjects:

ATH.png
When we speak of 20-20kHz in hifi, see how 1-dimensional this statement is? Because it is (ab)used by marketing to convince us to buy irrelevant expensive junk.
So when you speak of a high frequency roll-off of 3dB at 20kHz that is a 0-dimensional statement (not to say unsubstantiated or erroneous; it should be 3.9dB at 22kHz with no reference to how steep the roll-off is, 6? 12? 18? 24 dB/oct?).

I've never seen anything like this information before, I feel it is really helpful (to my) general understanding of how (human) hearing works, Frequency Response, SPL, and et. cetera., also how they are interrelated.

 

Thanks for sharing this, btw do you have source links, or links to similar information (I'd be interested in learning more). ?
 

 

post #813 of 935
Just basic googling of what I learned in school. If you know nothing Google is pretty useless. Just when you have some knowledge you can start shifting information from static.
This is just physics/acoustics/biology I learned over the years and progressive highschool/tech ed/univ.

What I didn't mention is that hearing is not only by your inner ear but also by touch, resonance of internal organs and hairs. The outer ear is needed for directional information and not just a flat horn. Especially lower bass you can feel with your gut. You can feel real bass. Thats why headphones can never convey the real feeling of being there.

And then there is protection of your hearing (healthy working environment). The hairs in your inner ear (cochlea) can only stand so much continued stimulation. The pain threshold in the graph you can only stand for seconds before nervecells at the base of the hairs in your cochlea start atrofying and you will have lost hearing specific frequencies for good being replaces by a constant beep or swooshing called tinnitus.
You can work in 70dB for 8 hours without damage, 80dB for 4 hours, 90 for 2h, 100 for 1h etc. Listening fatigue is your first warning, constant whistle your second and maybe last(ing) warning. That is why wearing headphones for extended time is dangerous even if it isn't that loud (you think). And so are concerts>100dB.

That should give you some more food for search.
post #814 of 935
Quote:
Originally Posted by ]eep View Post


Wow. Isn't that just a nice insult.
I'm merely giving information. You are not. You are just stating (unfounded) facts and opinions.
The writer conflates 'multi-bit DAC' with 'R2R DAC'.
So what is the difference? What else was there other than R2R before the invention of the sigma-delta DAC? If Philips invented the dac for CD what chips did they make before?
So then when he goes on to claim that Philips made R2R DACs, he's in error. They didn't.
So what are you saying? That the TDA1543 isn't made by Philips or that it is no R2R? Then what is it? Yes, a multibit, but how does it work?
Here is another site; same content (not surprising since they are in the test). enjoythemusic.com
It is clear that mother of tone is not the first with a NOS or ZO-DAC.
I would like to compare my modded Muse 4x with that competition. cool.gif
Burr Brown stateds for its PCM1704: SoundPlus™ 24-Bit, 96kHz BiCMOS Sign-Magnitude Digital-To-Analog Converter.
Here is the pdf where it states 24-bit resolution. and the noise-free you can see in the oscilloscope pictures: it still shows a sinus at -120dB compared to a flat line with a 20bit dac.
This is a quote from a developer (who seems to actually know what he is talking about and is willing to explain in plain English) that says it all imho (src).
Just what reality isn't linear? Man, you are so confusing the issues. How can a violin IRL not sound like a violin IRL? How can reality not be reality 1:1 in normal Newtonian physics?
Yes a loudspeaker has to be linear to make a recorded violin sound lifelike. And so does a headphone. For earplugs it is different however. There the acoustics of our outer ear have to be taken in account and this give a different frequency response. I think the Etymotics website has good information on that.


The insult was yours - 'arrogance' without providing evidence. No insult intended in anything I've written. This is just another example of your self-image being out of touch with what you actually write. You claim to be 'just giving information' but where was the information in the claim of 'arrogance' ? That was opinion for sure, and still unsupported opinion despite my request for you to show where the arrogance was taking place.

 

So, on to the technical issues. The difference between 'multi-bit' and 'R2R' type DACs is that there's more than one way to implement a multi-bit DAC. R2R is just one way - using a sequence of resistive attenuators. Another way is to use weighted current sources. Yet another is to use a resistor string. Philips has (to my knowledge) used the second option in its designs.

 

If you pull up the TDA1543 datasheet and examine the block diagram of the internals, you'll see there's a box marked 'current source' and this feeds into boxes marked 'dividers'. This means they take a reference current and repeatedly divide it in two to create the bit weights. This is their 'economy' DAC - on the TDA1541 the process is similar but higher accuracy using 'dynamic element matching' rather than passive dividers.

 

As to the PCM1704, on page 4 you'll note they show sinewaves at diminishing levels. The lines get fuzzier as the levels go down. That's noise.

 

Regarding linearity, I'll save that for when you've shown you've digested the above. Suffice to say at this stage that the transmission of sound waves through air is non-llinear because whilst there's the potential for almost infinite compression, there's a limit in the other direction beyond which we can't go - i.e. a vacuum.

 

post #815 of 935
Quote:
Originally Posted by Sapientiam View Post

As to the PCM1704, on page 4 you'll note they show sinewaves at diminishing levels. The lines get fuzzier as the levels go down. That's noise.

Regarding linearity, I'll save that for when you've shown you've digested the above. Suffice to say at this stage that the transmission of sound waves through air is non-llinear because whilst there's the potential for almost infinite compression, there's a limit in the other direction beyond which we can't go - i.e. a vacuum.

Your so full of ****, do you mind if I don't care to digest that? Although I'm very capable thank you very much.

You seem to know everything but never learned how to weigh (correlate) data.
The lines get fuzzier as the levels go down. That's noise.
Yes I see that that 'fuzziness' is noise. I'm not the idiot here. That 'noise' is 120dB down. Why are you bringing up 'noise' that is at least 30dB below the hearing threshold at maximum healthy listening levels while the signal is clearly visible? Noise is something that is loud and irritating to our ears in plain language. You are again confusing the issues here by using marketing psychology.
The noise you speak of is noise as in noise versus signal 'Physics: A disturbance, especially a random and persistent disturbance, that obscures or reduces the clarity of a signal'.
Not noise as in noise versus music or noise as in 'Sound or a sound that is loud, unpleasant, unexpected, or undesired.'
Here is another definition of noise: 'Computer Science: Irrelevant or meaningless data.' That sounds a lot like most of your contributions here.
Or this one: 'Remarks or actions intended to convey a specific impression or to attract attention:' If it's unwanted then stop making noise.
Or is that me again in just another example of my self-image being out of touch with what I actually write.? Or is it just you not being able to understand what I am saying?

Here another confusing example of lack of logic:
Suffice to say at this stage that the transmission of sound waves through air is non-llinear because whilst there's the potential for almost infinite compression, there's a limit in the other direction beyond which we can't go - i.e. a vacuum.
No, its omnidirectional, in 3 dimensions. Not linear, not 2-dimensional, 3-dimensional. And fades away in time caused by the friction of the medium.

I know the maximum soundpressurelevel at 1013hPa or 1 atmosphere is about 180dB (eq to a sonic blast of a bomb explosion) where pressure varies between 0 and 2026hPa. That literally changes like the weather and its a totally different story on Uranus (getting dangerously close to your **** again:rolleyes:). But what has that to do with non-linearity of the frequency-range or our accurate perception of sound? You are barking up the wrong axis. It affects all frequencies.
Or do you mean the harmonic distortion caused by the subsonic frequencies of a coming storm, or the doppler effect caused by it? Humans can't hear that. Elephants can however.

For the record: I don't agree with quite a few ideas on the 'mother of tone' website either. Especially the namesake theory. But it does have some clear insights that show a lot of thought and experimentation has gone in to it. But also a lot of filosophy that seems to want to go back to the magic era as opposed to science. I'm cool with against the grain, leftfield, out of the box original thinking but it should go hand in hand with a sound dose or relativity.
Whoever makes a bare PCB fitted on a spruce board (no paint but sticky lacker) for a DAC, phono-amp and amplifier and charges several €1000 for it? Doesn't he have a wife who has to look at it all day and dust it once a week? Or maybe he is blind that he has such acute hearing. I like my gear to look visually pleasing and my living room to be just that and not look like a laboratory or workshop.
post #816 of 935


 

Originally Posted by ]eep View Post

 

Quote:

Your so full of ****, do you mind if I don't care to digest that? Although I'm very capable thank you very much.

As you seem incapable of reasonable discussion without ad hominems, you're best left to your own ignorance. Pearls, swine.biggrin.gif

post #817 of 935

Hello There all

 

First post, so hope I'm not butting in here.

 

Am currently reading my way through this thread and not wanting to miss all the fun I've ordered one of these today which will be plugged into my Muse USB/SPDIF box (which I got for that reason).

 

Seeing as how a decent linear power supply seems to be a must the first thing to be addressed and being really cheap i realized I have more then enough bits in my various boxes to knock up a good regulated supply excepting only the power plug for the Muse itself. Can anyone tell me what size it uses so I can get one in ready? There are load of the little things and they all look much the same :-(

 

If this has already been answered somewhere, sorry, like I say I'm still reading through

 

All the best,

Graham

UK

post #818 of 935

My Muse arrived today, I've not plugged it in yet as I wanted to check I out first, but it's the latest white board version and has the correct 3.3v regulator in :-) I ordered the version with Elna Silmic II output caps in to save a job (I shan't be running it without output caps no matter what ones I end up with..

 

Been as this is the version with 12-18V printed on the board what's the thought on the power supply voltage? Should it still be 9v or is 12v OK now does anyone think?

 

Whatever happens I shall be building a new better quality supply, it has the tackiest least capable looking switch mode with it I've ever seen, doesn't exactly inspire with confidence even by switch mode standards!

 

Any thoughts would be welcome

 

Graham

post #819 of 935

I still reckon that 9V is a bit marginal and risks the regulator not giving its best ripple rejection. Either reduce the output voltage a little (to <7V with a resistor tweak) or run at least 10V input. Common-mode noise is the major issue with power supplies, more so than differential mode which regulators can filter out fairly well. This is what makes using an SMPSU a can of worms - it can be mitigated with liberal use of CM chokes on the input.

post #820 of 935

Thanks, I shall start at about 10V and play about a bit from there. However I just fired up the little beastie for the first time complete with dreadful switch mode supply and frankly I'm totally amazed. Right out of the box with no burn in it's showing my Rega a clean pair of heels!

 

Easy to listen to, the depth and seperation are stunning. It's easy for instance to pick out the backing 'strings' way down in the mix are in fact synth etc which I'd never noticed before, it's not in your face but it's there if you want it. God I hate to say it but it sounds analogue! NOW I can see what everyone's been raving about :-)

 

What will it do with a decent supply?!

 

 

post #821 of 935

Hi

 

please could someone wise tell me if buying this one DAC wouldn't be pointless for me?

 

I am using PC + creative soundblaster audigy 2 zs gold + creative speakers + kx project drivers (really nice sounding). I got mainly FLAC files + 24/96 files. So I would connect this DAC to spdif out (optical or coax) from creative soundcard front panel and to my little segments Technics SE-HD301 equipment (cinch). I am aware that this Technics system component is not something really cool or expensive, so my question is: does this setup not decrease DAC's signal quality?I mean amplifier - does the signal wouldn't be harmed? I am attaching pictures of block diagram and schematic diagram of this Technics (please open them in a new window to get the full size):

 

http://i210.photobucket.com/albums/bb167/tomahawk1981/blockdiagram.jpg

 

http://i210.photobucket.com/albums/bb167/tomahawk1981/schematicdiagram1.jpg

http://i210.photobucket.com/albums/bb167/tomahawk1981/schematicdiagram2.jpg

http://i210.photobucket.com/albums/bb167/tomahawk1981/schematicdiagram3.jpg
 

Thanks.


Edited by Heruka - 3/22/12 at 6:25am
post #822 of 935
Registered especially for this question?
Quote:
does this setup not decrease DAC's signal quality? I mean amplifier - does the signal wouldn't be harmed?
No. Why would it?

The sound is determined mostly by the weakest link. The DAC wouldn't be. The s/p-dif would do just fine if it can output 24/96. The amplifier is most likely the weakest link here, after the speakers of course. Upgrading the source is not a bad idea for this little money. For twice that you can throw in a nice t-amp too so that would put the burden of blame all on the speakers. You have to start somewhere, don't you?
post #823 of 935

Quote:

Originally Posted by ]eep View Post

The s/p-dif would do just fine if it can output 24/96. 


Yes, it can.

 

Thanks for reply anyway.

post #824 of 935

Linear stabilised DC power supplier @ 9V  800mA will be good enough for this dac?

 

thanks.

post #825 of 935
It's not to much but it will do. Which one are you talking about? The Teradac PS? You can adjust that to 10V.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Dedicated Source Components

Gear mentioned in this thread:

Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › Mini Dac TDA1543 X 4 NOS