PCM5102, 2VRMS ground centered direct output with two digital filters

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
fwiw, you don't actually play a flac directly; you decode (often into a memory 'pipe') and then play the literal waves.

if I decoded the flac to wav and played that, it would be the same experience for the listener. even a scope would agree ;)

flac is lossless compression that saves you about half over the raw .wav file.

anyway, let me cut out a section (its the intro, IIRC) and I'll upload to my public ftp.

I could ignore, to some degree, this problem if it was only on 'internet radio' as a source; but if its on fairly modern cd's, then this becomes a problem you can't just ignore.

I actually do hope its experimental error on my part. I would like having dollar-dacs as usable building blocks.
 
Yes, I have a laptop, but without SPDIF output.
It has analog audio out, but then I do not need the PCM5102.

In addition, my (one & only) PCM5102 is built with a 45.1584MHz XO which is divided by 4 to generate 11.2896MHz for the QA550 SD Card player.
The inputs are I2S. To work with my own WM8804 module, the XO would have had to be 48MHz.

And yes, I can build another complete new sets of modules to play your FLAC script.
But I rather thought that converting FLAC to 16 bit WAV would be somewhat less effort ?
Perhaps you can post a link to such a converter ?


Patrick
 
Last edited:
wait, I'm really confused.

how can you NOT get what flac is about??

its simply a format like .zip.

this talk about spdif and i2s is irrelevant. a .flac file should be playable by most modern software players and some pure hardware based ones. if you have a laptop, you can play flac. get foobar2000 (you should have that anyway if you do audio stuff on your pc) and find the flac plugin. that's what I recommend.

the slooppy way is to get the flac decoder, decode and waste space about 2x to go from flac to .wav and then play your wav's directly (or burn to cd, etc).

flac does not touch any audio content; and so if I rip 44.1k into flac, its still 44.1k when played back, bit for bit.

if I record an old DAT tape (48k) to my spdif soundcard and flac it, its still at 48k when the DAC see's it, at the output of spdif or even i2s.

dude, you need to get familiar with flac. its a very real (and useful) audio tool and you can't 'be in audio' and not know about flac. seriously.
 
Disabled Account
Joined 2002
Maybe that QA550 device Patrick uses won't play FLAC ? Just a hint....

Otherwise I agree with your post. FLAC (and APE) are good lossless formats to use. Way better than lossy MP3 and the like. Too bad most people use lossy formats and don't even hear the difference (they say). Even worse some devices are not able to play FLAC as they are MP3/WMA only :headbash:

Personally I don't hear any difference between FLAC and WAV which should be but I heard claims of the opposite.
 
Maybe that QA550 device Patrick uses won't play FLAC ? Just a hint...
and what's the problem? get the FLAC files and decode (convert) them back to wav before transferring 'em to the player!

There's no reason to waste time and network bandwidth and/or HDD space to transfer or archive wav files.

Lossless compression is just that, lossless. You can encode/decode and encode/decode again and again as many times as you want without loosing or altering a single bit.
 
Last edited:
i am reading to the pcm5102-Q1's datasheet,

• Integrated High-Performance Audio PLL With
BCK Reference To Generate SCK Internally

"System Clock PLL Mode
The system clock PLL mode allows designers to use a simple 3-wire I2S audio source then driving the DAC. This reduces the need for a high frequency SCK, making PCB layout easier, and reduces high frequency electromagnetic interference. The device starts up expecting an external SCK input, but if BCK and LRCK start correctly while SCK remains at
ground level for 16 successive LRCK periods, then the internal PLL will start, automatically generating an internal SCK from the BCK reference. In the PCM5102-Q1, the internal PLL is disabled when an external SCK is supplied; specific BCK rates are required to generate an appropriate master clock. describes the minimum and maximum BCK per LRCK for the integrated PLL to automatically generate an internal SCK. " [page 12]

The SCK can straight away tide to ground, but the question is will the chip itself generate a better SCK signal compare with the supplied SCK? Let say from TE7022L or TE8802L. Any comment on this?
 
Last edited:
The SCK can straight away tide to ground, but the question is will the chip itself generate a better SCK signal compare with the supplied SCK? Let say from TE7022L or TE8802L. Any comment on this?

I used the PCM5102 with a WM8805 tranceiver, which has better jitter specs than those Tenor chips. Nevertheless the result was not impressive. So I would use it with it a dedicated crystal. But if you are into quality, then try different DAC chips. The PCM5102 is very good because it's low cost and simple to implement (single supply, no master clock needed).
 
Anyone try the PCM5122 or 5142 yet ?

Nope, But based upon the datasheet, I think the analog part of those chips is exactly the same as the PCM5102. They have simple (PCM5122) or more advanced (PCM5142) DSP onboard. Thus you can do digital equalizing etc. For the price that is very interesting. And it is much better documented than the DSP functions of the ESS Sabre chips.
 
Just browsing trough the datasheet of the PCM5102, and there is a interesting point:

SCK rates that are not common to standard audio clocks, between 1MHz and 50MHz, are only supported in software mode, available only in the PCM512x and PCM514x devices, by configuring various PLL and clock-divider registers. Software mode allows the device to become a clock master and drive the host serial port with LRCK and BCK, from a non-audio related clock; for example, using 12MHz to generate 44.1kHz (LRCK) and 2.8224MHz (BCK).

Thus, the PCM5122 and PCM5142 have a built in PLL that can generate a MSCLK. Because it is not asynchronic, as with Sabre dacs, it will probably not improve upon a good 256fs external master clock. But it saves you switching between clocks for 44.1/48KHz sources.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.