CD16 reclock problem

[Isaac Sibson]

Hi everyone.

I have been working on my Marantz CD16, adding a trichord digital output board and power supplies. Here's a pic of the work in progress:



I'm going to shorten the clock leads when I mount and connect the DOB outputs.

However, at the moment I have a rather odd problem arising. The player works fine when first powered up, but within 15 minutes the DAC (modified M-Audio Superdac) starts losing data lock intermittently and then completely, and then the transport mechanism goes haywire. Switching the player off and then on again sorts it out for the next 15 minutes.

Any ideas as to what the problem is? Could it be injected noise on the power supply (as I've seen some 33.8688MHz carrying along the DOB power supply connections)?

Any help gratefully received!

[jean-paul]

At what pin of the NPC digital filter did you connect the new clock line ? Please at least twist the 2 wires or use flexible coax at that spot.

Compliments for the tidy way you've built things in the cdplayer !

[Isaac Sibson]

I didn't. The new clock is connected to the Xin pin of the TDK 7WU04 resonator chip. Should I connect it somewhere else?

[jean-paul]

Yes, check the schematic and connect the clock line directly to the right pin of the NPC chip. I don't know CD16 or its schematic well but this is the procedure that gives best results in general.

[Isaac Sibson]

I've done that now, pin 6 of the 5803.

Now the transport mech spins continually (even with no disc inserted), and won't read the TOC.

[Isaac Sibson]

Here's the latest situation:

[martin clark]

The transport spinning out of control is usually a sign that there's something wrong with the clock feed to the decoder (the SAA7350 here).

Check your black lead is connected to a digital ground point common to both the DAC and decoder digital grounds.

Do essentially the same for the red lead: trace the clock routing to both the DAC and the decoder chips, and see if they are fed by a common inverter. Then either run two new feeds (twisted pairs) from the trichord to the correct (Xin and digital ground) pins on both chips; alternatively feed your new clock signal just to the inverter on the board doing the distribution -probably the more reliable option, in case there is anything trick in the Marantz implementation about relative timings of the clock signals.

If you run your new clock in direct, make ure you cut the traces for anything else 'upstream' so that existing inverters do not load-down or fight the new clock feed.

Good luck!
M.

[Isaac Sibson]

Ok, made a bit of progress. It's now reading TOC again, having moved the clock input to pin 3 of the left hand HC74 below the 5803. This comes from the output of the 7WU04 (pins 2 and 3) to which I formerly had the clock set as the input. Let's see if it holds up over 15-20 mins though...

[Isaac Sibson]

Nope, it now loses the plot after about 1 minute.

[Cobra2]

- but not home before wednesday...

Arne K