New DAC measurement

[Sylvania]

Hi

Dac datasheet.. Price guess it is priceless.

How did they remove the digital noise in the teac player?

Best regards

Kim Olesen

 

[jives11]

Will post soon PCM1702, PCM1702-J, Playstation 1 = SCPH-1002, and maybe *** ***** S2.

PCM1702 is very poor, PCM1702-J is better, Playstation 1 is like 14 bit TDA1540.

Hi Bernhard,

any chance os posting the Playstation SCPH1002 results, eiither here or in the "playstation as a CD player" thread in digital. Would be interesting to see how it looks

Thanks

 

[Bernhard]

Playstation... ok.

I have this TEAC CD700 player with 1 x PCM53 that is so good.
Now got a CD880 with same digital noise system but 2 x PCM 54 and it is very bad. Perhaps it is the PCM54 that is a crappy chip in general. Never saw a good one.

 

[Bernhard]

Hi

Dac datasheet.. Price guess it is priceless.

How did they remove the digital noise in the teac player?

As far as I remember it is something like alternating noise - noise &signal - noise with sample-hold and inverting = cancelling.

Not very helpful to know about because IMHO impossible to implement with DIY tools :-(
You need the TEAC chip.

 

[Bernhard]

Hi Kim,

about the PCM56 in the Teac player: It was so good because Teac added digital noise and removed it in the analog section.
So the DAC never see very small signal and distortion is always low.

Sorry, did a typo here , of course it is PCM53.

PCM53 does not seem to be that good by itself, it is the TEAC chip doing wonders.

I have a Sony CDP502ESII with 2 x PCM53-K and it is average in measurement and sounding a little bit Yellow.

No MSB adjust pot implemented tough. ( on my to do list )

To do list:

ökjhfbrkgäzhft#öp#lz
ertkgbfndlznädotfät
detngdätnetgndeo
kljfelertfkjerjfkotkdciuh
irehwheugkjgbtirojir
eoifgjru,kjeripiotrtit0trt
wiergouetrioh8tirkf3rgu
pouweorfirujkde39u8oi
poeurhfiuofrirfro3ughoirf
oierfrroigjrfioirijfrwtrorwghoi
poruiooferrotioiruhioirwzuhfei
irurfuijfiuhriofjigrgiorjiogrojougi
oieufiw^trjifrgugrfiuiereihfoeruih
eeruhfdfeijfoerfuieuioefueuofeieu9fi
eifirfuoiuoieuvigueiwr89fe0wiodhfu8u9re
oiuegfiuzfoe8fe87ueif8uief8uhoie98ueiie9uzo
dfewoife8oidefufieiwf8uieiwrf8uidoieq9udfddo
eudefueirzufeiqrfde0rf9uhijoweif9sdufdioeifvfuvi0fe9fu
9ufifv8eusdieszduieddie89fdhuiewf98udieifdhijie9dhife0i
ddf3euidfoi3uif0siofiw89sudfifef8uohfie98uivfe98oifef9sdi
uddi0dfvidifuiods0afsuhideife9sdu9eq8zdf9e09fduudi0dua
uudfasiuduew8asdiudewuasdudadsv uvie9us8uvfi9afdsuuad
dsudadsifueia9fsudduauigudeazudu9fzv guduzuhdud8zuidu98zu
uu9asfhdeuauihdiuedzuhdfuszvuvifuasz8uvvdhidasvhdusvuhdudfgfgh...

...

 

[Bernhard]

Sony Playstation SCPH1002

Noise floor -55dB

K3 -40,4dB

All very bad, but sounds sweet, soft & lazy

 

[Bernhard]

Above Picture, whole graph was only shifted 10dB , for direct comparison again correct 20dB and with cursor on 1kHz:

 

[davidsrsb]

[Q]How did they remove the digital noise in the teac player?[/Q]

Sounds like they added a known low level dither signal to the pcm and the subtracted it again after the DAC. To diy this you would need a dsp or fpga to do the maths and a second DAC.

This is only really useful for a multibit DAC and reduces harmonics at the cost of raising the noise floor.

 

[Sylvania]

Hi

Looks like a simple solution ( fpga and an extra chip). I think i will look at ebay for an old player with a PCM53, and make an external DAC... or find another chip type.

 

[Bernhard]

[Q]How did they remove the digital noise in the teac player?[/Q]

Sounds like they added a known low level dither signal to the pcm and the subtracted it again after the DAC. To diy this you would need a dsp or fpga to do the maths and a second DAC.

This is only really useful for a multibit DAC and reduces harmonics at the cost of raising the noise floor.

It more reduces the harmonics than it raises the noise floor.

A second DAC will introduce problems because it has a different harmonic spectrum. Two DACs of same type are almost never the same. Sample/hold like in Teac should be easier.

 

[Bernhard]

Hi

Looks like a simple solution ( fpga and an extra chip). I think i will look at ebay for an old player with a PCM53, and make an external DAC... or find another chip type.

Teac still uses that kind of chips in modern playern even VRDS25X IMHO.

Somebody posted it sounds better with the chip bypassed
But that can vary from model to model, the smaller model CD700 is better than the bigger CD880...

 

[Sylvania]

It more reduces the harmonics than it raises the noise floor.

Add / Sub data before and after the ADC, will move the data to a more linear part of adc. The ADC will then make less Harmonics because of less distortion.

Sample/hold like in Teac should be easier

what i need is an external spdif/dsp/adc solution... but first project is adding a SPDIF output on the player.

 

[Bernhard]

Add / Sub data before and after the ADC, will move the data to a more linear part of adc. The ADC will then make less Harmonics because of less distortion.



what i need is an external spdif/dsp/adc solution... but first project is adding a SPDIF output on the player.

The noise is added permanently so the DAC will never see low level data and never produce the harmonics.

 

[davidsrsb]

The dither signal will be only a few bits so the cancellation signal to the secondary DAC can be multiplied by 256 and the output scaled. The result is negligible harmonics from the secondary DAC in the output.

 

[tube-lover]

Hi Bernhard,

I would like to test for the NPC SM5864.
This also a 20Bit DAC.
I got some stock for this DAC with SM5842AP/5843AIS. Digital filter. I will combine with AD1890JN to prepare for my another DAC.


Do U ad interest to test. pls email to me for your address & I will post two pcs for U to test.

I hope to collect same advise/comment for this DAC from U.

thx

thomas

 

[Bernhard]

Hi Thomas,

thanks for offer but I already have a pair, removed them from a board of an Onkyo player.
SM5864 are delta sigma and Onkyo used TDA1547 in the top models.
If I will get a complete player somewhere someday, I will test the SM5864, but build a DAC is too much work, I am only interested in multibit for my own project.

 

[tube-lover]

Hi berhard,

I will get the selected 2002 versio taiwan TDA1541a chips, but I had no way to test. Would U care help me to test. I can post one pair for U to test. I hope to know what is the quality of the new made chips. The hearing result was smooth & little bit detail than the Old TDA1541a. And.......any comment for the SM5864!!


thx

thomas

 

[Bernhard]

If you want I can test the 1541A but the results will only show what you might expect from the rest because every single chip is different.

SM5864: No idea, only know they are bitstream and maybe not worth to make a board.
Why you choose them ?
I can post you mine for evaluation samples.

What about AD1862 ? I have a pair waiting to be tested but no time yet.

PCM56 is good chip also

 

[tube-lover]

Hi berhard,

I chose SM5864 due to the source.

I was designed the new DAC which was 20Bit.

USE CS8414>>AD1890(16MxTCXO)>>SM5842AP(with TCXO) than Pass to SM5864.Due to I seperate several PCB. cab only adjust the SM5864 & the TCXO than User can exchange the DAC.

AD1862 was quite expensive if combine with AD1890, SM5842AP in one DAC.

any comment for my decision??.

pls advise. Your comment was very important because I know U like multi-bit.

thx

thomas

 

[Bernhard]

No idea about the SM, the datasheet does say 20 bit input word but what does that say.
Why not make a balanced PCM1704 board ? People want that.