Thule Audio's GND Management by Integrated Amplifier with RC - who can explain

There are two grounds: Analog GND and digital GND (D-GND)
The kind of creation I don't understand and I have never seen by others.

At first look it shows like a symmetrical supply voltage about +/-5VDC.
But there is only one voltage: +5Vdig.

Because no additional secundary winding for independent creating of the +5Vdig is in use, this way was choiced from Anders Thule.
The neg regulator drives a bidirectional diode (normally anti paralleling 1N4148) and creates 0V7DC refer to GND (A-GND). But the actually GND wire for the neg regulator goes here to +5Vdig !!
What about this idea from Anders Thule?

Who can give me a commonly circuit describtion ?
 

Attachments

  • Thule Audio Bal-ONE GND Man.ckt.pdf
    9 KB · Views: 279
Last edited:
There are two grounds: Analog GND and digital GND (D-GND)
The kind of creation I don't understand and I have never seen by others.

At first look it shows like a symmetrical supply voltage about +/-5VDC.
But there is only one voltage: +5Vdig.

Because no additional secundary winding for independent creating of the +5Vdig is in use, this way was choiced from Anders Thule.
The neg regulator drives a bidirectional diode (normally anti paralleling 1N4148) and creates 0V7DC refer to GND (A-GND). But the actually GND wire for the neg regulator goes here to +5Vdig !!
What about this idea from Anders Thule?

Who can give me a commonly circuit describtion ?
Any news ?
 
So by the looks of it (to me). The positive regulator creates the +5V then, as its reference is that same +5V, the LM337 generates another 0V or GND linked with the original by the paralleled diodes. The diodes are there for protection so the grounds do no move far apart. In my opinion, it's not a good circuit. With the tolerance of the resistors and the parts you could end up with a reasonable voltage difference between the grounds. The circuit isn't that necessary anyway. I've designed FPGA DSPs with 24bit audio DACs on the same ground and PSU and achieved over 120dB noise floor with proper PCB layout.
 
So by the looks of it (to me). The positive regulator creates the +5V then, as its reference is that same +5V, the LM337 generates another 0V or GND linked with the original by the paralleled diodes. The diodes are there for protection so the grounds do no move far apart. In my opinion, it's not a good circuit. With the tolerance of the resistors and the parts you could end up with a reasonable voltage difference between the grounds. The circuit isn't that necessary anyway. I've designed FPGA DSPs with 24bit audio DACs on the same ground and PSU and achieved over 120dB noise floor with proper PCB layout.
yes. best solution for integrated amplifiers of Anders Thule's "SPIRIT" series is to create an outdoor power supply with independent transformers for +5V dig., +/-7V5 for volume control, +/-15V for the operational amps so as a mono power supply for each channel of power amp section. A lot of work - OTOH best solution and no longer thermal stress on the PCB around the SO-8 voltage regulators.
For reasons of cost, one absolutely wanted to avoid independent transformer windings - therefore Thule's genuine approach with diodes, hot resistors and burned PCB that make unhappy.
 
Last edited: