MK2703 PLL clock in SACD, good or no good?

Status
Not open for further replies.
Jump to Latest
tubee

tubee

Member
Joined 2004
#1
I have a Yamaha S657 SACD/DVD/CD upsampling player. Have downloaded the service manual to look at the schematics. (philips DVP 720 is allmost the same)

The clock is provided by a PLL chip, the Audio Clock Synthesiser MK2703. Is this any good?

Another SACD player, the Philips DVP 963SA has 3 cristals which are switched (But switching of clock signals is not the best way imo)

What can possibly done to mod the clock of this Yamaha SACD player? I only can think of good bypassing with sanyo oscon near PLL chip.
 
#2
tubee said:
I have a Yamaha S657 SACD/DVD/CD upsampling player. Have downloaded the service manual to look at the schematics. (philips DVP 720 is allmost the same)

The clock is provided by a PLL chip, the Audio Clock Synthesiser MK2703. Is this any good?

Another SACD player, the Philips DVP 963SA has 3 cristals which are switched (But switching of clock signals is not the best way imo)

What can possibly done to mod the clock of this Yamaha SACD player? I only can think of good bypassing with sanyo oscon near PLL chip.
Click to expand...

it provides the correct frequencies (good), but jitter wise..........

Neverthjeless, these PLL chips provide better clocks if their source is better: Replace the standard crystal attached to these ICs with a better clock and you will bennefit. Decoupling helps, but better take a separate regulator for the PLL chip.

best
 
#4
Hi Guido

I was hoping allready that you could reply to this question, thanks a lot.

Didn't open the player, but saw indeed in the datasheet of PLL a cristal is used, the block diagram of Philips DVP720 didn't mention about a cristal. I thought the frequency was defined by the PLL chip only. Will upgrad clock to PLL, and build separate (& choked) PS for it.

Hi Jon.

Yamaha S657 has used same Philips mainboard of Philips DVP720, but omitted the eco board and has another keyboard pcb. I could only find 720's serv. manual.
 
#5
The PLL uses MCL from mechanism, and buffers it. No cristal near PLL used, but there must be a cristal found somewhere, i assume near DVD mechanism.
Opamps used: LM833, yuk. Should try LM4562's
 

Attachments

  • schermafdruk-1.png
    schermafdruk-1.png
    89.2 KB · Views: 356
#7
tubee said:
The PLL uses MCL from mechanism, and buffers it. No cristal near PLL used, but there must be a cristal found somewhere, i assume near DVD mechanism.
Opamps used: LM833, yuk. Should try LM4562's
Click to expand...


mcl from mechanism, yuck....

I wouldn't spend too much money, other than to see if replacing the master clock on the mechanism, improves anything

best

Guido
 
#8
I wouldn't spend too much money, other than to see if replacing the master clock on the mechanism, improves anything
Click to expand...

Thanks for tips Guido.
But this player doesn't sound bad at all out of the box. The cristal freq. must be tapped somewhere else, that's problably why they used a PLL chip afterall.
 
#9
I couldn't hold myself any longer, lifted the hood of the Yamaha. 3 pcb's, switching PS, mainboard and mecanism board. Opamps are 4 smd LM833,s
One cristal: 27000 ceramic smd(27Mhz) near mechanism.
What i could do is build a new and simple oscillator, buffer it with 3 And gates and lead the 3 clock signals to the parts who need it.

And what about this very simple clock: (found it on diyaudio somewhere else)
 

Attachments

  • simple clock2.jpg
    simple clock2.jpg
    20.4 KB · Views: 309
#12
tubee said:
I thought also it was a textbook design Guido.

Btw the PS for it is designed by C.Snoeren, he designed this originally for an Tent XO! Have a copy of it somewhere in my files. (original schematic removed from www)
Click to expand...

Hi

Yes, Cuno made this supply circuit. it performes well, but isn't ultmate in terms of low noise (with the Tentlabs supplies we are 40dB down compared to the 431 based supply)

best
 
#16
tubee said:
Will try your XO again with a "flea" on perfboard.

I want to buffer the proper clock with something and get 3 outputs, don't know with what chip yet. Maybe a simple 74HC04 will do the job.

Found something too:
Click to expand...

The XO can drive 3 HCMOS gates, no buffering required

the link says

" For the same clock application, the PLL synthesizer clock offers the use of a less costly crystal that can operate at lower harmonic frequency compared to a XO module. Replacing an XO module with a PLL synthesizer could shorten lead times and reduce BOMs. "

It doesn't tell yo that you get more jitter for free

best
 
#19
Hi Guido

I hope you don't mind, but have a Kwak PCB for it ready, but this one uses a plus-minus PS, not sure if i can find a +/- supply in the player. Your XO needs only a humble single ps. .😉
Kwak is stated for 25Mhz max, i need 27Mhz :cannotbe: but i probly think it should work even on 27Mhz.

Have to check around with scope if freq. is everywhere the same, and skip the PLL. Find another scope then first, have a 20Mhz max.

For better opamps four LM4562's, and coupling cap the small & cheap BG C types are planned. In future maybe a separate PS and transformer for the opamps (now one general switching PS)

Personally i think a analytic sounding XO will match good with the original LM833 opamps, for the LM4562's i don't know really.
 
#20
A flea in my furr!

I have a flea ready, made on perfboard in cheap version (NE5534 instead of AD797) Mounted a Tent XO 11.2896 Mhz on it. Use it on a CD660, PCM56 non-os with I2S glue logic out of SAA7210 and Pedja's AD844 I/V stage.
It all looks now like a bomb wich has to be dismantled😉

The I/V has a separate 50VA ringcore transformer, the flea/Tent a small separate transformer. The CD660 is modded in PS with new caps and BYV diodes, CDM2 modded too (lead, rubbersheet & woodglue/sand in clamp)

Sound? Cool!😀
 
Status
Not open for further replies.
Top Bottom