Using Ground Plane on Both sides of Board?

[ppl]

Has any one had experience using a ground plane on both the component side as well as the foil side of a two layer PC Board? I use ground planes on the component side often with good results and every so often get tempted to put a ground plane on the Bottom side also so as to enclose each foil with a shield. Conventional wisdom suggests that two ground planes arranged with one above the other as would be the case with putting a ground plane on each side of the same PC board would induce capacitive coupling between the planes causing undesired circuit operation. By placing vias thought the board so as to short the top and bottom plane together could perhaps reduce the capacitance coupling to the point that the undesired circuit operation would not occur for Audio amplifier circuits.

Your thoughts please

 

[jan.didden]

ppl,

Assuming that both planes are actually gnd planes, there is no voltage between them therefore no cap effect.
But in reality there may be tiny (hf) voltage differences between areas so stitching them together or using the component holes where possible to tightly interconnect them seems a good idea.

I always try to have gnd planes on both sides and haven't seen any negative effects.

However, I always do my high current and power supply tracking separate and isolated from the gnd planes just to avoid potential differences across the plane. I use the planes only for small signal returns (input and output returns) and screening.

Jan Didden

 

[PMA]

Better to use 4-layer PCB. Proper layout is quite difficult. Paralleling ground planes may bring an issue - HF resonator effect. Sometimes it may be better to use just one ground plane, compared to 1st for GND and 2nd for +Vs.

 

[ppl]

Thanks for your insights. yes the power is on its own plane and conected to the main plane only at one point,However small current bypasses for the voltage gain stage is to the main plane.

 

[peranders]

As said it's better to strive for one single and also massive groundplane unless you don't wan't some extra cooling like I did in my QRV-07 design. To be honest I haven't checked how much extra this did. It certainly did not make it worse.

Paralleling ground planes may bring an issue - HF resonator effect.

I think for audio, this is nothing to worry about.

Solder side
Component side

 

[PMA]

From strictly 20kHz point of view, surely not. Unfortunately, supply rail issues are one of the main factors influencing the perceived sound.

 

[peranders]

I think you must "think" up to at least 1 MHz, possibly more.

 

[darkfenriz]

Per
Have you ever heard your cellular (900MHz or 1,8GHz) noise in (audio frequency) speakers?

 

[peranders]

Yes, but not in my own gear

 

[jan.didden]

Yes, but not in my own gear

Indeed. There are lots of incompetent designs out there in this respect. So far, my own designs are not part of it. It really takes not rocket science to avoid GSM breakthrough. But if you built a complex amp from just a circuit diagram and no prior experience, I think you are lucky if you get it to work at all, unless you buy a kit and even then it's hard.

Jan Didden

 

[darkfenriz]

I once lied an old nokia cell phone directly on the PCB of 100W amplifier, no shields, no plane grounds. The noise was 'just hearable' when I called this phone on 86dB speakers.
Do you therefore mean that PCB was that bad?

 

[jan.didden]

I once lied an old nokia cell phone directly on the PCB of 100W amplifier, no shields, no plane grounds. The noise was 'just hearable' when I called this phone on 86dB speakers.
Do you therefore mean that PCB was that bad?

No, of course not! This is an extreme situation. Like building an amp that is 100% lightning proof. It probably can be done, but if your amp dies in a direct lightning hit doesn't mean it was a bad amp.

Jan Didden