Marantz CD63 & CD67 mods list

[Glenn2]

Originally posted by SimontY
Just been trying to remove the oscillation in my power supplies. I removed the silly 82n and 33n caps I'd put in, leaving just 1u for bypass at the psu end. Even after a couple of tracks, when I pause, there is no whine, and noise is fairly low. I think this has cracked it.
Simon

Good news!

Were those caps on the regulator output? Some regs don't like low ESR caps right on their outputs and will oscillate. LM317 can do this it seems. The output inductance creates a resonance with the cap. Should not be a problem with 78xx I think. Something to consider.

Also, most regulators need >10mA output current to work properly. If one is feeding, say, the DACs analogue stage or clock gates they only draw a few mA so you need to strap a <500R resistor across the reg's output.

[Malefoda]

For those who want to konw how much work has to be according to Ray's list, even if many parts are missing ( some removed caps, you'll have to add some surface caps to bypass some other and more... ).

And merry Xmas folks !!

[SimontY]

Originally posted by Glenn2
Good news!

Were those caps on the regulator output? Some regs don't like low ESR caps right on their outputs and will oscillate. LM317 can do this it seems. The output inductance creates a resonance with the cap. Should not be a problem with 78xx I think. Something to consider.

Also, most regulators need >10mA output current to work properly. If one is feeding, say, the DACs analogue stage or clock gates they only draw a few mA so you need to strap a <500R resistor across the reg's output.

Interesting thoughts... The caps were on the psu boards before the regulator boards. The regs are not feeding any big, low ESR caps. The regs are Audiocom super regs.

One reg feeds the clock divider chip (can't be many mA) and the other the Audiocom clock (also can't be many mA).

Perhaps the inductance of the umbilical wiring (1m+ including wire inside the CD) is playing with the caps in the power supply.

Some oscillation kicked in eventually, but it's much less severe than before, and it took about 10mins to start! So further refinements are needed

Simon

[martin clark]

Just where are your regs mounted w.r.t the loads they feed Simon? If you have them some distance away, then that could do it. Also, with anything less than the minimum distance you can arrange between the regs ground reference and the loads 0v pin, you're not getting the regulation you've paid for!

There's a lot to be said for sticking a largish, low quality (high -esr) cap at the output of a 3-pin reg (10-100uF), to damp any tendency to oscillate:
http://www.tnt-audio.com/clinica/reg...edance1_e.html

[SimontY]

Originally posted by martin clark
Just where are your regs mounted w.r.t the loads they feed Simon? If you have them some distance away, then that could do it. Also, with anything less than the minimum distance you can arrange between the regs ground reference and the loads 0v pin, you're not getting the regulation you've paid for!

There's a lot to be said for sticking a largish, low quality (high -esr) cap at the output of a 3-pin reg (10-100uF), to damp any tendency to oscillate:
http://www.tnt-audio.com/clinica/reg...edance1_e.html

Hi Martin,

The regs are fairly close to the ccts they feed, just a few cm of wiring away. There's room for improvement though, I'm looking at getting them even better connected.

It would seem to me that the part of the circuit before the regs is oscillating, not the actual regs. I do not really know! I'd like to make it better.

Simon

[6h5c]

Originally posted by Malefoda
For those who want to konw how much work has to be according to Ray's list, even if many parts are missing ( some removed caps, you'll have to add some surface caps to bypass some other and more... ).

And merry Xmas folks !!

Looks familiar

Ray

[marcusdeming]

Originally posted by markk02474
Marcus,

From your picture I see a couple things:
1. You didn't jumper the resistors powering the opamps. Its good to do, perhaps after listening with the HDAM removed so you can hear each difference.
2. I don't see a wire to carry the signal around the HDAM. I hope it is on the underside.

Hi Mark,

I have jumper the power resistors and also replace the ecaps at that area with some cerafines470uf/25v. The opamps, Lm4562 have arrived and are now working very well in my cdp.

I have also removed the caps btw the HDAM though it might not be relevant to do so. Is there any other thing that i need to jumper but miss out?

I am thinking of removing the headphone out next followed by some ecaps replacing in the DAC area and the clock hack mentioned in the accoustica.org (but i am quite unsure about what its saying)

What do you guys think will reap the most improvement and i should try out?

Regards,
Marcus

[marcusdeming]

Originally posted by 6h5c


Looks familiar

Ray

Hi Ray,

I see that alot of the orange colour bobs are removed. Are they the ones that are at the opamps area? What can we expect from removing them?

[SimontY]

Originally posted by marcusdeming


Hi Ray,

I see that alot of the orange colour bobs are removed. Are they the ones that are at the opamps area? What can we expect from removing them?

Those orange ones are ceramic. They do not belong in the signal path. Replace with silvered mica, polystyrene, teflon or polypropylene. You can expect a big improvement lol

If yours is SE or KI you won't have those in the output filter, they'll be mylar.

Simon

[Glenn2]

Originally posted by SimontY

It would seem to me that the part of the circuit before the regs is oscillating, not the actual regs. I do not really know! I'd like to make it better.

Hi Simon,

Are you pre-regulating at all?

Also, I've heard that the Audiocom super-regs are not unconditionally stable.

Do you have any capacitance at their inputs?
Also, sometimes we need that few cm of PCB track to dampen the reg's ouput. It's a minefield mate!

Martin will probably correct me (hope so anyway) but I think the input cap on a reg can be as low ESR as you like, and the cap bypassing the load device also can also be, with the caveat that the damping is provided in-between by the PCB track, and the cap on the reg's output is not too low ESR to cause ringing or oscillation.

LM317s can do this though I think you have to be quite unlucky, and I think low-dropout (LDO) regs are prone too, as are some super-reg designs (with remote sensing of load ground).

I think the bog-standard 78xx is only likely to oscillate with if your bypass caps are too high ESR (but not sure about that).

Are they all Audiocom jobbies or do you have others?

Also, could your noises be caused by ground loops?

I think there should be a simple solution to this....