Marantz CD63 - v.high level on one channel

[yestertech]

I am looking at one of the above players. Whilst the mechanism seems fine and the LH O/P plays normally, there is an odd problem with the RH audio O/P. The level seems to have increased about 10 fold - The signal is just very loud and extremely distorted. It's not moderate imbalance you might get with component drift or even reasonable gain variation.It's reminiscent of the sort of gross overload you'd hear if a line level input was plugged into a mic channel.(VV loud and distorted) Makes me think of untamed open loop gain or somesuch.
I can't find service info for this elderly unit.
Any ideas ?


Andy

[Radio Wrangler]

DACs most usually give a current output.

To turn this into a voltage, the current can be fed into a resistor to ground and Ohms law does the rest. Open circuit the resistor and you get too much voltage and the DAC output hits its voltage limits.

That doesn't give as much voltage as is usual in CD players, so instead of a simple resistor, an opamp is used.

The + input of the opamp is grounded and the DAC pours its current into the - input.

The opamp has far too much gain, it needs feedback. So there is a resistor from the opamp output back to its - input.

Opamp inputs are high impedance, so the only way for the DAC current to go is through this feedback resistor. The opamp needs to try to make the + and - inputs have the same voltage, so it swings its output to pull the DAC current through that feedback resistor until the - input is brought to zero volts. So the same Ohms law business happens, but with voltage inversion. Also the DAC output is driving its current into a virtual ground which helps its linearity. The gain available from the opamp allows the resistor to be larger, so more output voltage can be made. As if these advantages weren't enough, the opamp also buffers the output and its feedback gives it a good low output impedance.

So what could possibly goo wrong

Well, the DAC can't easily make the output of a faulty channel go a lot larger because they already are run to as much current swing as they can during loud passages. There is an outside possibility that thte DAC reference current system, that provides a DC reference setting the current scale of the DAC is bad, but the same system usually does both channels.

This brings us to the opamp circuit. A cracked/open circuit feedback resistor would fit the symptoms, maybe with a dry joint, maybe a cracked track. A good visual inspection around this area might turn up trumps.

David

[Edward Huggins]

The full service manual for the CD63 can be downloaded for free from

www.hifisounds.co.uk
http://www.hifisounds.co.uk/manuals/..._SE_KI_doc.pdf

A most excellent Philips built player - there was also an SE "tweaked" version.

[Sideband]

Originally Posted by Radio Wrangler

This brings us to the opamp circuit. A cracked/open circuit feedback resistor would fit the symptoms, maybe with a dry joint, maybe a cracked track. A good visual inspection around this area might turn up trumps.

David

We often found that the chip components could develop hairline cracks across the solder points...the chip itself was fine and just wanted the (briefest) of touches with a soldering iron. You might need a magnifier to see it. All sorts of strange effects (including distortion) could be cured in this way.

[mhennessy]

That's not a full manual.

Try this one: https://www.vintageshifi.com/reperto...ice-Manual.pdf

And of course, it's also on Hi-Fi Engine.

With that, it's standard fault-finding - plenty to get your teeth into! Hope it's not one of the HDAM modules, though they are only unity gain. Well, unity gain in electrical terms, but they offer huge amounts of gain in the marketing department! As you can see from the manual, if faulty they can be removed and bypassed by installing R619/R620.

The DAC is an SM5872 - datasheet here: https://pdf1.alldatasheet.com/datash...PC/SM5872.html - output is in differential PWM form, rather than current. The analogue circuitry is as straightforward as it gets

[mhennessy]

Originally Posted by Sideband

We often found that the chip components could develop hairline cracks across the solder points...the chip itself was fine and just wanted the (briefest) of touches with a soldering iron. You might need a magnifier to see it. All sorts of strange effects (including distortion) could be cured in this way.

These are mostly through-hole, especially in the post-DAC area, so hopefully straightforward - with any luck just a dry joint that should be readily visible with the right light and magnification.

Checking the manual reminded me that the HDAM "modules" are actually just on the main PCB, with a fancy screening can mounted over the circuitry. There is one SMT device in each (the 2SK332 dual JFET acting as the LTP), so that'll be worth a look if the fault if localised to that part of the circuit, although it's in a leaded package (SOT23, probably, IIRC), so not prone to that exact problem. But still could be suffering from regular dry joints, which is far from unheard of in these machines.

[percival007]

The 'original' CD-63 was a very different unit from the 63MKII
Which version do you have?

https://www.hifiengine.com/manual_li...tz/cd-63.shtml

or

https://www.hifiengine.com/manual_li...z/cd-632.shtml

[yestertech]

This one looks like the pic in the first link - alas no service manual listed ...

Thanks to everyone for the suggestions...
I've not had a chance to crack it open yet but the damaged PCB track sounds feasible. As suggested, it sounds as if a feedback loop is open, but there's a heck of a lot of gain being tamed !

A.

[mhennessy]

In which case, you'll need the manual here: https://www.hifiengine.com/manual_li...ps/cd101.shtml

I do have one of these on the to-do pile, and a copy of the original service manual somewhere. Electrically, the analogue sections are even easier than the 1990s version.

[knobtwiddler]

On top of the excellent anwers you've had so far, I'd add that the laziest / quickest way to fix it could be to locate the block that acts as the I-V amplifier and simply go around all of the pins with an iron and fresh solder. If you have a vacuum desoldering machine, you could remove the old solder first (a classic problem with 70s / 80s gear is ring cracks / brittle solder around large joints, going open-circuit). If you have said machine, you could desolder / re-solder the whole block in 5 minutes. If it doesn't work, then you've only wasted 5 minutes. Zig zag around between each pin so as not to heat each component, letting it cool before you do another leg.

[mhennessy]

Yes, that "special" Philips solder

It's been a while so I can't quite remember for sure, but if I do recall correctly, these machines don't use those brass rivets that caused so many problems in the CD104 and similar machines. So, that's one less headache, hopefully!

This machine uses separate DAC chips for each channel, and they are usually socketed, so if there's any doubts about those, it's easy to swap them between channels. But either way, while it's apart, it'll be worth re-seating those and the other socketed ICs. Those sockets aren't the worst, but I have had occasional intermittent problems with them.

[percival007]

Originally Posted by mhennessy

Yes, that "special" Philips solder

It's been a while so I can't quite remember for sure, but if I do recall correctly, these machines don't use those brass rivets that caused so many problems in the CD104 and similar machines. So, that's one less headache, hopefully!

This machine uses separate DAC chips for each channel, and they are usually socketed, so if there's any doubts about those, it's easy to swap them between channels. But either way, while it's apart, it'll be worth re-seating those and the other socketed ICs. Those sockets aren't the worst, but I have had occasional intermittent problems with them.

This unit uses the TDA1540 DAC chip, if swapping, please make sure you are Anti-Static protected, they are quite old now and may be a little fragile now. They're also possbly in the original ceramic package and becoming rarer than the proverbial hen's teeth in that guise.

A nice unit you have there, and quite valuable now I suspect.

[yestertech]

SO, I ve now had a chance to delve into this unit.
I'm grateful for the pointers to the service info for the Philips unit which is the same 'neath the covers.
I've attached a relevant snip of the circuit diagram which deals with the DAC and audio outputs.
The RH channel is high so I've checked around IC 6525 ( NE5532)
Ohms readings on the relevant pins of this ( measured on the IC itself ) revealed no obvious problem with the gain/feedback arrangements -
1K8 between pins 2 and 1 ( first stage )
direct link between pins 7 and 6 ( second stage )
supply rails correctly connected to the same points on the NE5532 on the LH channel.

Which leaves me wondering if the IC 6520 ( TDA 1540 ) could be the culprit, but then where does the extra gain come from ?
The NE5532 is not socketed , but could it be worth replacing this first ?
I'm loath to swap over the TDA1540's for fear of damage, but I guess this could be the only way of proving whether the fault stays or moves ?

A.

[mhennessy]

I'd be amazed if it's the 5532. Nothing is impossible, but they are generally very reliable IME.

I think I would begin by assuming that IC6520 is doing it's job correctly - especially as you have measured 1k8 between pins 1 and 2 - and question if the DAC is producing too much current (unlike in the later model, the op-amp in this case is indeed being an I to V converter, as David explained in post #2).

You can check the current reference resistors (connected to pins 15 and 17 of the DAC), but after that, I'd just swap the chips around. Normal static precautions should be fine - I'm not aware that they are especially fragile in that respect.

[PJL]

If you have a scope you can compare the L and R channels signals and work backwards from the output to locate the fault.

[knobtwiddler]

The only time I've seen 5532s (or 5534s) die has been when I've killed them

[percival007]

You could try desoldering the O/P Pins on both TDA1540's (Pin 22) and 'swapping' where they go with a couple of wires.
Left iOut to Right Op-Amp and vice-versa.

This should tell you if it's it the DAC or the circuitry afterwards.

Just be careful with static and don't leave the iron on the pins too long. If they're socketed then remove the IC's whilst doing the de-solderig, placing them on a static free surface.

[percival007]

Originally Posted by PJL

If you have a scope you can compare the L and R channels signals and work backwards from the output to locate the fault.

Unfortunately being Current Out DAC IC's you can't see anything on the 'scope.

[mhennessy]

In the examples I've seen, the TDA1540s were socketed...

[PJL]

Originally Posted by percival007

Unfortunately being Current Out DAC IC's you can't see anything on the 'scope.

Not sure I understand as the current is converted to a voltage in circuit. There are then 2 op-amp stages which can also be checked by comparing with the other channel.