Accurate <100 ppm HV49 crystal??

[fmak]

Does anone know of a source for a 7.3728 MHz HC49 crystal with a guaranteed <100 ppm or better frequency.

I am modifying a satellite receiver to output a spdif stream which will be locked by a two pll relocker.

Or do I have to but an XO with additional power supply?

[JohnW]

FMAK,

I don't know the Sat Box your modding, but 7.3728MHz sounds like a common Frequency for Toshiba system MCU (I use a Toshiba MCU in my CD player / Transport servo board designs) - are you sure its related to the Audio Clock – or are you trying to Sync clocks to reduce system noise.

I might have so other suggestions if you explain the system,

I would have thought that all Xtals will have greater then 100ppm accuracy - or is there a ceramic resonator in your box?

I can send you a HC49 7.3728MHz crystal if it helps.

BR,

John

[davidsrsb]

7.3728 is a standard data comms value, 64 x 115,200bps

[fmak]

[QUOTE]Originally posted by JohnW
[B]FMAK,

I don't know the Sat Box your modding, but 7.3728MHz sounds like a common Frequency

I might have so other suggestions if you explain the system,

---------------------------------------------------------------------------------
I am using a Sony VTX 670 sky digibox to access uk radio at 48kHz in Cyprus. The unit uses a CS4922 chip which outputs pcm via toslink. Already the sound is much better than DAB in England using a Sony STD777ES. With the cheaper DABs the sound is diabolical!

To improve the sound, the toslink output is fed into an Assembalge D2D dejitterer with twin plls. The secong pll will only lock within 200ppm or so but is not working with this unit.

The CS4944 is fed by a 7.7328MHz crystal via a HC74U04 chip and I am assuming that the clock frequency is out.

This is from an inspection of the unit from the top; no dissembly yet.

Eventually I shall add a spdif output from the tx pin. Got any good circuit?

If you accept emails, I shall send you my address for one of the crystals.

Grateful for your comment.


Fred

[JohnW]

Hi Fred,

I expect that once you take a better look inside your box, maybe under the PCB, you will find another crystal, the 7.3728 MHz dos not sound like the correct part for the Digital audio section - although I could be wrong.

If there is not a second crystal for the Audio Clock, then I expect that the PLL that generates the Audio Rate clock from the 7.3728MHz is not exact.... changing the crystal will not help... I expect you will find that the PLL jumps between discrete "offset" frequency’s to generate an "Averaged" 48 KHz output, this is why the 200ppm VCXO circuit will not lock.

Attached is a similar output from a DVD decoder chipset - note the "Discrete" changes in Frequency to generate an average "48KHz" clock from the 27MHz video master Clock.

I would not be surprised if the audio clock was not generated by the 27MHz MPEG Video clock (if there’s one in this box) - just as a within DVD player.

As I said, I don't have any idea about this Sat Box (or anyother Sat Box for that matter!), so this is just an "educated guess",

Hunt down that 27Mhz Crystal, and you will have your answer,

John

[fmak]

If there is not a second crystal for the Audio Clock, then I expect that the PLL that generates the Audio Rate clock from the 7.3728MHz is not exact.... changing the crystal will not help... I expect you will find that the PLL jumps between discrete "offset" frequency’s to generate an "Averaged" 48 KHz output, this is why the 200ppm VCXO circuit will not lock.

------------------------------------------------------------------------------------
John

You are right, although I still cannot find the 27 MHz crystal - underneath?

As an easy step I measured the frequency at the clkin pin, 27.0005MHz; 48kHz ouput probably jumps as you say.

How do I get round this?

Fred

[JohnW]

Fred,

Sorry for my very poor English in my earlier post - a bit of reverse logic there - but you seem to have understood what I was trying to say...

It's would now seem most likely that the Audio Master Clock is being generated from the 27MHz MPEG clock by a poor PLL circuit - if the PLL is poorly designed within the Decoder IC there’s not much you can do.

Within 6 months or so, I can offer you a "Jitter Buster" box that will solve your problem - but you will have to wait until then....

John

[JohnW]

Fred,

After some thought, I believe the 27MHz MPEG Video clock must be VCXO based to allow synchronization with the Satellite broadcast – I expect you will find the 27MHz VCXO in one of the RF Screened Tin Boxes – if there are any…

Does not help much – as you cannot improve upon this Clock – as it needs to Track the broadcast Frequency….

John

[kevinkr]

Are you sure that the Assemblage device supports sample rates other than 44.1kHz? It might be redbook cd compliant only.. (Some other reclockers I have run across that use a vcxo lock only on 44.1kHz)

Kevin

[fmak]

[QUOTE]Originally posted by kevinkr
[B]Are you sure that the Assemblage device supports sample rates other than 44.1kHz? It might be redbook cd compliant only.. (Some other reclockers I have run across that use a vcxo lock only on 44.1kHz)

------------------------------------------------------------------------------------
Of course 24/96 otherwise the first pll won't lock!