Positive power supply rail at a different potential than negative rail?

[The Lord Flashheart]

Further to my poking around inside a Sony XM 6020 I have discovered that the positive rail potential is ~33% lower than the negative rail's potential to ground.

The positive rail measures at around +22 v whereas the negative rail measures at -32 v.

I have noticed that the filter capacitor on the negative rail is also warmer than the capacitor on the positive rail.

The spec given in the service manual is +32/-32 volts.

The voltages at the drains of the mosfets just before the transformer are identical.

The square wave taken from the anodes of the rectifiers seems to be "offset" low, if that makes sense?

Is the problem in the rectifier diodes, it seems unlikely to be the transformer isn't is?

 

[Bigun]

that power supply schematic suggests that the rails should be symmetrical and hence either a) somewhere excess current draw is pulling it down or b) rectifier / caps are not doing their job. If you measure across L901 you might get some idea of how much current is returning through the gnd, if the amp is symmetrical it should be very little if there's a lot of current there then something is pulling the rail down.

 

[Osvaldo de Banfield]

Some of the MOSFET's get hotter than others? One of the arms may be faulty, or the gate resistor altered.

 

[The Lord Flashheart]

that power supply schematic suggests that the rails should be symmetrical and hence either a) somewhere excess current draw is pulling it down or b) rectifier / caps are not doing their job. If you measure across L901 you might get some idea of how much current is returning through the gnd, if the amp is symmetrical it should be very little if there's a lot of current there then something is pulling the rail down.

I'm sorry but I can't find an Inductor marked "L901" on the schematic. There are two inductors in the output stage marked "L101" and "L201", have I missed something?

 

[The Lord Flashheart]

Some of the MOSFET's get hotter than others? One of the arms may be faulty, or the gate resistor altered.

The mosfets on the (+) rail in question are stone cold (running at idle with no input signal) whereas the mosfets on the chanel with the normal (-)rail voltage are just ever so slighgtly warmer.


The gate resistors look orginal, as does everything else in this amp, I can measure them in circuit to see if they are out of spec.


Should I do this with the power off or on?

 

[Osvaldo de Banfield]

With the set powered off.

Do you have an oscillo available, and are you able to measure gate signals? Different temperatures make me thing that one of the two parallel or both are, or overloaded or misdrived.

 

[The Lord Flashheart]

Yes, I posted the gate signals in another post:


Bottom is gate and top is drain. All measure the same in shape and magnitude.


I have also found that there is a potential of around -4.8 v between the battery negative input and the centre terminal of the transformer on the rectifier side.


The gate resistors measure within tolerance.

 

[The Lord Flashheart]

I'm sorry but I can't find an Inductor marked "L901" on the schematic. There are two inductors in the output stage marked "L101" and "L201", have I missed something?

Edit:


I did miss it, apologies Bigun.




It is glued to the board however and I cannot make any measurements without taking the board off the heat sinks, it does not seem to get particularly hot however.

 

[Osvaldo de Banfield]

Check L901, between the 'lytics. May be open circuited.

 

[The Lord Flashheart]

Check L901, between the 'lytics. May be open circuited.

It is not possible to measure it's resistance directly from the top of the board but measuring between the transfomer centre connection and "blu" and "brn" on the speaker output board gives a resistance of 0.5 ohms.


That's low but is it a short?

 

[Osvaldo de Banfield]

10µHy is a very low value and then, few turns of copper wire.
The rails once rectified, looks equal, measured not to -bat , but to speaker ground?

 

[The Lord Flashheart]

If I have read the schematic correctly, the speaker ground are pins 1&3 on the speaker board, then the potentials are +- 26v as measured on the collectors of the output transistors.


Does that mean it's fine?

 

[The Lord Flashheart]

Thank you for your continuing help by the way Osvaldo, very much appreciated.

 

[Osvaldo de Banfield]

If between collectors and amp ground both rails measure equal, the no fault.

Note that battery - and circuit ground is separated by a resistor of 2.2KΩ (or so) and a 0.1µF cap (C912 and R909), so, current flowing in this R alters your measurement.

 

[The Lord Flashheart]

Sadly not, not between speaker ground and the collectors.

 

[Osvaldo de Banfield]

But in post #12 you measured bot equal?, I'm getting lost :-(

 

[Perry Babin]

Re-check R909 as was suggested. Do this with nothing connected to the amp. You should read 2.7k between the primary ground and the non-bridging speaker terminals.

Do you have a signal source plugged in while making these measurements?

 

[Osvaldo de Banfield]

Thank you for your continuing help by the way Osvaldo, very much appreciated.

Don't worry, this is why I'm here.

I was thinking where are the measurement made. The oscillo and the voltmeter, respectively, if to bat ground or amp ground.

In other hands, I don't understand why to separate both systems, useless for me.

 

[Perry Babin]

If you don't isolate primary and secondary grounds you will get 'engine noise' due to ground loops. This configuration is common in car amps. An alternative is to have the primary and secondary directly connected and use a balanced input on the signal input circuit.

 

[Osvaldo de Banfield]

OK, thanks for teach to me. I have little experience in such a car amplifiers, and sincerely, I dislike them per se. But in the few devices I repaired, both rails was only one.