Dual power suuply PCB

[dr.power]

Plz can you guys kindly answer my post # 15 too?

 

[dr.power]

I noticed that puting 100 or 200uF caps on the output of the regulators will surely shortens the regulators for a small amount of time. Isnt it harmfull?

 

[Brevor]

Plz can you guys kindly answer my post # 15 too?

You should not have any problem with ground loops.

 

[dr.power]

You should not have any problem with ground loops.

Thanks. But why are you thinking so while I think the ground CURRENt in the pic of post #15 has Two ways to reaches to the output ground or vice versa? Actuall I think that It is better to disconnect the outer ground loop from the ground and let it to be free (not connected to anywhere). Any idea plz?

 

[Boncuk]

Thanks. But why are you thinking so while I think the ground CURRENt in the pic of post #15 has Two ways to reaches to the output ground or vice versa? Actuall I think that It is better to disconnect the outer ground loop from the ground and let it to be free (not connected to anywhere). Any idea plz?

The red rectangle of the PCB design as in post #15 is obviously drawn on the component layer, which means the PCB must be double sided.
If you do use a double sided PCB just call the red rectangle "ground" as you should have done on the solder side. Both grounds are connected physically already via the ground pins of the connectors.

If a 100nF cap on the output creates a short the 100nF cap on the input would also create one.

That's nonsense since a cap won't let pass DC - it will only be charged.

Just build the circuit as planned and rearrange the parts for a complete ground pour for the entire board.

Boncuk

 

[dr.power]

The red rectangle of the PCB design as in post #15 is obviously drawn on the component layer, which means the PCB must be double sided.
If you do use a double sided PCB just call the red rectangle "ground" as you should have done on the solder side. Both grounds are connected physically already via the ground pins of the connectors.

If a 100nF cap on the output creates a short the 100nF cap on the input would also create one.

That's nonsense since a cap won't let pass DC - it will only be charged.

Just build the circuit as planned and rearrange the parts for a complete ground pour for the entire board.

Boncuk

Opps! Actually the red rectangle is not something more than an INDICATOR, I wanted to show the ground loop portion, so I used an read color indicator (the board is just signal layer).

We did not talk sbout 100nF ceramic caps as the element of the shorting, we meant the big electrolyte ones at the out put (for instance a 100 or 200uF cap). is it make sense plz?

 

[davenn]

Opps! Actually the red rectangle is not something more than an INDICATOR, I wanted to show the ground loop portion, so I used an read color indicator (the board is just signal layer).

We did not talk sbout 100nF ceramic caps as the element of the shorting, we meant the big electrolyte ones at the out put (for instance a 100 or 200uF cap). is it make sense plz?

he prob typo'ed maybe should be....
If a 100uF cap on the output creates a short the 100nF cap on the input would also create one.

but regardless the output cap should only be ~ 1 - 10uF electro and preferably along with a 0.1uF
On the input to the regulator you should have ~ 1000uF electro if you are planning on current up to the 1Amp max of the 7805 (7905) and again a 0.1uF
the general rule of thumb is 1000uF / Amp. 2200uF wont hurt, but if you used 1000uF you would be able to more easily mount them on your board

Yeah there's been much debate over the years about using the 0.1uF 's on the ins and outs of the regulators. You may well get away without using them
but then... hey... the chip designers wouldnt have included them in practical designs without reason and for their tiny price its cheap insurance again any
possible oscillations. In low freq audio work probably not really critical. BUT for the RF work I do, I wouldn't dream of power rails without lots of decoupling
to overcome the probability of RF causing PSU modulation issues.

All caps should be mounted as close as possible to the input and output pins of the regulator chip(s)
your redrawing looked much better

Dave