DIYHiFi.org

This site
http://en.hifidiy.net/DIY_Production/20 ... 9_2358.htm
says remove the black part. With pliers?
Anybody has an idea how I can get out the SMD crystal?

BTW what 's pro about this oscillator, being part of the SAA7324?

Hi Elso,
They recommend 'shaving' off the crystal (and breaking the internals) because it is difficult to lift off the board unless you can't heat the four pads simultaneously. We actually did just that because the result is 'cleaner'.

Here is the link to our modified CD-Pro.

I was lucky enough to have access to two of these hot tweezers so for me it was an easy one man job. In case you can't find one of these useful tools (good for soldering/desoldering almost anything smd up to MSSOP8 size) you could always ask a friend to bring in his 'classic' toys...just make sure you have suitable clean tips and that the lenses are protected from scratches, not exposed excessively to fumes...etc.

For the second part of your question: nothing Pro about that, just an ordinary CMOS clock contained in the decoder and exposed to all 'jitters'.

Hi, Jocko suggested Chip Quik.
I will give that a try.Thanks.

On second thought the clocking scheme of the SAA7324 is more complicated. For 8.x or 16.x input clocks there is a pll multiplier (x8 or x4) that generates internally the real MCLK of the chip. For 33.x this pll is disabled (pin 24 - SELPLL must be tied low) and in theory it should give the best results...

But if you reclock everything it shouldn't be much of an issue, irrespective of frequency.
As far as I remmember 16.x cannot be used without modifying some configuration register so only 8.x and 33.x are "diy-friendly".

I only see a 8.4672 MHz crystal oscillator which I will replace by my own clock....

Here's how I'd do it...

Cut two pieces of bare wire (a resistor leg would work well) that are the length of the crystal leg. Lay one of them across the two leads on one side of the crystal, and solder it in place with a big blob of solder that flows into each lead, and across the wire. Repeat on the other side. You should have a nice big blob of solder on each side of the crystal.

Now, get a second soldering iron, and a helper with a pair of tweezers. Melt both sides simultaneously - you'll see the crystal lift up a bit when everything's melted and flows, and when this happens, get your helper to lift the crystal out of there.

Clean up the pads with solder wick. Done.

Once the xtal and the small cap on the input side are removed you can take advantage of the pads 2 and 3 of the (former) xtal to solder a 1206 series resistor or coupling cap...depends on what you intend.

I don't intend to bug you with such obvious suggestions, just sharing from my 'vast' experience with the CD-Pro

I finally decided to abort the clock exchange in the Proceed CDD transport.

I saw a wire coming from the crystal going all the way over the main board to a CY2292 IC. Also close to that is a can type oscillator part# Madrigal 810006.
The owners manual states:
"Using a custom-made crystal oscillator with better
than five part-per-million accuracy, the digital signal is reclocked immediately be-
fore the output, eliminating transport-related jitter from the digital audio signal.
This same crystal oscillator controls the all-digital servo used to control the rate
at which the disc spins."

Well, well the CY2292 is not really a low jitter device. In absence of any schematic I gave up.
I cleaned and lubbed the drawer rails.
And also after building all together again I ran into an awkward drawer problem.
On closing it opens halfway so playing is impossible.
After trying all possible resets unsuccessfull I got really mad and jammed the tray in. Then I let it it play 10 minutes which appeared to be normal.
After that open and close of the drawer was normal.
I dunno what caused it, finneky software??

I do not recommend this transport. My Proceed PCD3 with my own clock sounds less "flat".