Voltage Regulator Ground Question

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

CJ

Well-known member
GDIY Supporter
Joined
Jun 3, 2004
Messages
15,664
Location
California
input power ground does not get connected to the output ground

(red circle)  that input ground goes to the xfmr CT.

they are creating a virtual ground with the transistors?  thanks for any help!  :D
 

Attachments

  • reg.png
    reg.png
    479.1 KB · Views: 69
Thanks John! (your better with a beer buzz than i am sober,  :D  )

had dis similar voltages on left and right channels after recap job, checked gain of all transistors and resistor values ,

had a cap in backwards (red circle) not in a place where it would blow up, but in a place to cause skewed voltages,

this looks to be a load dependent design so rails should come up after install into Sansui stereo.

 

Attachments

  • Sansui AU717 pwr sup v.png
    Sansui AU717 pwr sup v.png
    641.9 KB · Views: 29
Hi CJ
Powering up the PCB ASSY on the bench you need to jumper the grounds. Once the "i #22) connects to Ground the regulators will work. The Ground connection is made through other circuits on the other PCB's.

If the jumper is placed inside the "red circle" there will be some ground currents that will increase the "HUM LEVEL"
Duke
 
That is not that uncommon when a subassembly has an open ground path, that gets completed by the rest of the assembly.  This is often a strategy to manage ground integrity (referencing it elsewhere.)

This can be problematic for development and production testing at subassembly PC board level, or repairs.

It is not crazy to just design in a modest resistor (say a couple hundred ohms) to complete the ground path for testing, but still look like a relative open circuit when assembled into the full system. 

JR
 
Audio1Man said:
Powering up the PCB ASSY on the bench you need to jumper the grounds. Once the "i #22) connects to Ground the regulators will work. The Ground connection is made through other circuits on the other PCB's.

If the jumper is placed inside the "red circle" there will be some ground currents that will increase the "HUM LEVEL"

Casual eyeballing told me this as well, but good to have confirmation.

A thing of note is that not using a hierarchical ground and resorting to this kind of unusual configuration reeks of PCB layout and design issues. And they leaked all the way to the schematic which is not ideal.
 
Kingston said:
A thing of note is that not using a hierarchical ground and resorting to this kind of unusual configuration reeks of PCB layout and design issues. And they leaked all the way to the schematic which is not ideal.
But it is a hierarchical layout. Pin 04 of the card is the filter cap ground. Whatever that plugs into probably has a bus of points connecting to various things like the chassis, earth ground, other cards. Its the "star" in star grounding. It seems potentially superior to me actually because the regulators just need a sense of what ground is but you want high currents to return directly to the filter caps (preferably alongside supply lines).
 
working good now,

replaced 2SC1708  NPN 55 volt with 2SC1845  120 volt and

2SA847 PNP 55 volt with 2SA992  120 volt

voltages are with minimum input to get 33 out, Veb of TR10 moves up from 0.347 on up into conduction zone as input rails are brought up,
 

Attachments

  • Sansui AU717 pwr sup voltage.png
    Sansui AU717 pwr sup voltage.png
    604.8 KB · Views: 9
squarewave said:
It seems potentially superior to me actually because the regulators just need a sense of what ground is but you want high currents to return directly to the filter caps (preferably alongside supply lines).

If those supply currents have a signal component, especially a half wave rectified signal component,. one could also argue that they should be sourced, terminated and hopefully cancelled local to the circuits that generate and consume them, and not all the way back to the supply. Having rectified signal current flying around a large console is a recipe for muddy sound when you have a lot of channels summed to mix.
 
Monte McGuire said:
If those supply currents have a signal component, especially a half wave rectified signal component,. one could also argue that they should be sourced, terminated and hopefully cancelled local to the circuits that generate and consume them, and not all the way back to the supply.
The major half wave rectified current is returned via a direct path to the transformer CT.  The clean side of the ground break only sees ripple voltage through a RC filter.
Having rectified signal current flying around a large console is a recipe for muddy sound when you have a lot of channels summed to mix.
There are numerous concerns when summing a large number of signals.

JR
 
Back
Top