Voltage Regulator Problem?

[fmak]

Sony SCDXB 940 SACD player. The digital filter is fed from the analogue +- 12V regulators via +-7V and 3.3V regulators.

I am trying to remove clock related pulses from the analogue power supply which finds it way into the output via OPA2604 filter/buffer. The stock unit is littered with inductors and capacitors to provide isolation of each power supply stage.

Having tried lots to further bypass opamps and regulators, I am wondering why there are 560 Ohm resistors connecting the inputs and outputs of the +-7V MSF 7X07 regultors and whether this is the problem. These regulators are fed from the analogue ones and isolated by 30 uH inductors at their inputs.

Can anyone wiser than me explain the purpose of the resistors and why they are used only for this specific set of regulators please. I am intending to remove them once I know why.

[jam]

Could you post schematic?

Jam

[Jocko Homo]

Are you sure that your 'scope leads aren't picking these up?

Jocko

[fmak]

Originally posted by Jocko Homo
Are you sure that your 'scope leads aren't picking these up?

Jocko

Yes, I first noticed the problem when measuring unweighted THD which was only -80 dB below reference; high for a cd player. Display on scope direct show sabout 30 mV pk-pk of nasty high frequency spikes corresponding to clock freq. Weighted, the unit meets spec. I suspect the sonic signature to be related to the noise; hence the desire for a cure. The same pattern obtains in the analogue power supply rails, the filter/buffer output and can be made much worse by further decoupling between the V+ and V- rail of the filter/buffer opamp.

Another possibility is that the opamps themselves are the cause. They are OPA2604s with 20 MHz bandwidth. I have come across this problem of roque opamps in the SCD777ES but SMT mounting means that I don't want to remove and refit unless I have to.

I can't post schematics as I don't have a scanner. The issue is:

Why does anyone want to connect input and output of 78 07 and 7907 with a small SMT resistor of 560 Ohm with a voltage differential of 5V.
The unit used appears to be 1% film as well!

[Elso Kwak]

Hi fmak,
These resistors are a silly habit of Sony. Silly beacause they provide a path for spikes and other garbage from the unregulated supply to enter the regulated supply. I had one such resistor too in the +5V digital supply of my Sony CDP. I removed the resistor and replaced the regulator by LT1086. Much better.[not pin compatible!] Maybe they have done it as the regulator gets overheated otherwise. Another silly Sony habit: inadequate cooling (too small heatsink) Hope this helps as I owe you one!

[fmak]

Originally posted by Elso Kwak
Hi fmak,
These resistors are a silly habit of Sony. Silly beacause they provide a path for spikes and other garbage from the unregulated supply to enter the regulated supply. I had one such resistor too in the +5V digital supply of my Sony CDP. I removed the resistor and replaced the regulator by LT1086. Much better.[not pin compatible!] Maybe they have done it as the regulator gets overheated otherwise. Another silly Sony habit: inadequate cooling (too small heatsink) Hope this helps as I owe you one!

Can't be overheating problem as the resistor will not take any appreciable current. Must be an other 'technical' reason????

[fmak]

[QUOTE]Originally posted by fmak
[B]

I removed the input /output resistors and this has no effect. I am beginning to suspect instability in OPA2604. Has anyone found a similar problem. The noise spikes are low level high frequency but look very nasty.

[PassFan]

There are some older audio amateur articles from the 90s that dealt with noise in digital equipment power supplies as well as some mods and super regulators. These are great for reference if you plan on going deep into digital and can be found at www.audioxpress.com in their backissue section.

[hagtech]

Elso is right. Designers often use resistors in parallel with series regulators to either help dissipate the heat or keep the maximum load current within the limits of the regulator. It's cheap. Sometimes pennies count.

The resistors do not really add a path for spikes. Combined with the output bpass capacitors an RC low pass filter is created. Very little of the RF gets through. Of course, it depends on the quality of capacitors.

The best way to eliminate clock pulses & such is to use ferrite beads. They work better than inductors at RF. On the other hand, Jack is probably right - it's your scope probes causing the problem.

jh

[fmak]

[QUOTE]Originally posted by hagtech
[B]Elso is right. Designers often use resistors in parallel with series regulators to either help dissipate the heat or keep the maximum load current within the limits of the regulator. It's cheap. Sometimes pennies count.

The resistors do not really add a path for spikes. Combined with the output bpass capacitors an RC low pass filter is created. Very little of the RF gets through. Of course, it depends on the quality of capacitors.

The best way to eliminate clock pulses & such is to use ferrite beads. They work better than inductors at RF. On the other hand, Jack is probably right - it's your scope probes causing the problem.

------------------------------------------------------------------------------
Thanks for the help. If it is the leads, how do I get rid of the pickup. I have tried ferrites on these and inductors on the output of the opamp without success. Needless I have re-routed cable runs etc. This will be the first time I have come across the problem on any CD or SACD player. The spikes appear whether the top is on or off with no change in level. As I said, thid unit has more inductors and bypassess than I have come across normally.