Philps CD960 bad sound

[Cleanharry]

Recently bought myself a nice second hand CD960 CD player. At first the sound was very impressive. After having listened to it for several weeks I feel there's something wrong with how the CD960 sounds. It can be described as follows: lots and lots of details, deep but no strong bass and a strong emphasis on transients. It sounds a bit loud in the higher mids, not really harsh and a bit untidy every now and then. Now again some brighter recorded discs are hard to listen to. What puzzles me is that when used as a transport and driving my DACMAGIC DA converter, the impression more or less stays the same: different character but a bit of roughness, a bit untidy. Very different from my Philips CD850mk2 when driving the DACMAGIC. Then the sound is very polite and even slow. I didn't expect a transport having that much influence on the sound quality.

Since I feel that both analog and digital output of my CD960 suffer for this bit of roughness I suspect to be something wrong (or wore out) in the disc reading and/or decoding chain.

Any suggestions out there?

btw: the CD960 (CDM1 drive) reads every disc I throw in. My CD850mk2 (CDM4/23) even reads CD-RW's.

[Thomo]

The digital filter chip (saa7220) really needs a high-quality power supply with its own voltage regulator. This will improve all areas of the sound, including digital output.

Of course, clocking will help enormously too.

Lee.

[SoNic_real_one]

IMO SAA7220/TDA1541 does that anyway - on analog out. I have a CD473 and is the same. SAA7220 it's not such a great filter, regardless what fans of it say. And TDA1541 without it (in NOS mode) doesn't sound better either.

Also, the circuit that drives the spindle motor is important, because it can generate lots of low-frequency jitter that is not eliminated by the 16kB buffer memory in the decoder SAA.

The fact that you say that sounds (almost) the same via SPDIF it is probably related to the internal generated jitter of the SAA and beefing/separating power supply might help. I didn't try that - after replacing the opamps and adding capacitors on power lines and output, I got bored of it.

[UV101]

Hi, you player history is similar to mine. I ran an 850MKII with a Trichord clock 4 for many years before messing with several CD63's and then landing in the TDA1541 camp.

To that end, I also have a philips CD960 which is virtually identical to the Marantz CD94MKI. What I can tell you is that the std sound is quite very poor like most of the early 1541 players. This player is better than most (as it should be) as it has a substatial PSU and discrete requlation for the output stage.

The CD960 is capable of sounding truely stunning for sure, but it takes some work. The absolute minimum to make this player start to sing, is to use very good regulation on the digital filter chip SAA7220p/A located on the vertically mounted PCB to the right.

Another thing to consider is that every electrolytic cap in the player is over 20years old. There is not a single original cap left in my own player this will be effecting the sound for sure. Replacement with carefully selected caps will result in audio superior to the original sound.

I would start by replacing all PSU smoothing caps and all local decoupling caps. I'd be suprised if this did not result in a significant performance increase. If you want advice on this process, you are in the right place

Edit...

There's a few pics of the various stages of modification here. You can see the replaced caps etc...
HifiSounds CD960

[mace1337]

Drop in a separate Salas shunt reg for the filter chip.

[UV101]

Originally Posted by mace1337

Drop in a separate Salas shunt reg for the filter chip.

I did!!! Then I took it out again and used a decent linear reg!!

[mace1337]

Originally Posted by UV101

I did!!! Then I took it out again and used a decent linear reg!!

Hah. Carry on then.

[UV101]

Seriously, there were a lot of changes made when I went to the shunt regs and something totally killed the sound. One of the reasons I went back to the linear regs was that I wasnt unhappy with them to start, I just wanted to try something new. The linear regs I use are easy to get inside the player and the shunts were massive. With all the PSU mods, there was no way they'd fit inside the player.

The fact still remains that the filter really needs a decent independant reg.

Another great thing about this set up is that you can directly clock the decoder, filter and DAC if you use a clock that generates /2 & /4 of the master clock. Then you dont have a daisy chained clock with each device in line adding jitter.....works very well.

[Cleanharry]

Thanks a lot for your swift responses.

It seems that for the time being I have to concentrate on 3 issues:
1. poor regulation on the digital filter chip SAA7220p/A
2. worn PSU smoothing caps and local decoupling caps
3. worn caps in motor spindle and laser circuits

And what to expect from a new low jitter clock circuit: great improvement or discussable?

Btw. it seems that working on this CD960 is quite easy since most parts are easely accessible.

[mace1337]

This could work quite well:

1PPM 11.2896 MHz Low Jitter TCXO Crystal Clock for DAC | eBay

It has the divider that UV101 talked about.

You'll need to check if the 11.2896 is correct for your player.