Small distance from ground, what will happens?

[destroyer X]

I normally call my adviser related to those doubts...he is travelling or busy to answer, this way i will open one question... a little bit ashamet related ignorance.

Lets suppose you assemble components in a turned up side down board.... i do this way... since i was a kid.

So, no conduction of eletrons because phenolic and fiber glass are very good insulators.

But, i glue them to ground metal chassis.... so i think capacitance will be created because off two pieces of metal 1 mm distance (may be 2 milimeters or something alike)... in audio amplifier,
may i expect some trouble?
Is it a good way to avoid oscilations?
Will this induce oscilations?

I suppose 1 picofarads, or half each point... can be helpfull or a problem.

Carlos

 

[Jay]

Good question destroyer X. I often took similar approach intuitively, but didn't think if it should be a problem. It is not a plate (or is it something else?) it is just copper path. And shielding signal path with ground path is a good practice.

 

[destroyer X]

I really do not know

And sometimes, when ashamed i go straigth into a friend... he is too much busy.
This way i put here, because we have very special knowledge people that normally do not like to show themselves.... but always a chance to wake them up to help me!

Hugh from Aksa
Rodd is another

And there are more 10 i suppose that i could not discover them.

Carlos

 

[azira]

But, i glue them to ground metal chassis.... so i think capacitance will be created because off two pieces of metal 1 mm distance (may be 2 milimeters or something alike)...

Well, firstly, you are basically creating a capacitor from EVERY trace to GND.

So, thinking on the high side, lets say it's all hand soldered work, no PCB etching. In this way, I'll assume your traces are 3mm wide. We'll also use 1mm distance to Metal since that is the worse case number.

C = eo*A/d eo = permativity of air = 8.85 x 10 -12 F/m. A = area in meters^1, d = distance between plates in meters.

So, we can start with a particular length, say 1cm, and get a characteristic capacitance for every cm of trace which can be multiplied out for each trace.

So for 1cm of 3mm trace.

C = eo * .01*.003 / .001 = 2.655E-13. That is .27 pF.

1pf for each 4cm of trace about.
--
Danny

 

[ChocoHolic]

Hi Carlos, hi Azira,

please note that the capacitance will be higher, because the dielcetrical material is not air but phenolic.
The relative dielectricity constant of phenolic or bakelit ranges between
3.5 ....5 from what I know. So factor 4 may be a reasonable approximation.

Bye
Markus

 

[destroyer X]

Hello Azira and Chocoholic

Thank you...but nobody told me if this can be a problem or not.

This way i suppose so small capacitance will not make it oscilate because will ressonate in SHF frequencies.... and those transistors do not have response to oscilate... this way, this capacitance represents nothing?

Am i rigth.... that's what i was thinking, but full of doubts because i learn all by myself, this way... a lot of doubts, i cannot make those calculations and have those values to think about.

thank you all

Carlos are thinking that's no problem!

best regards

Carlos

 

[ChocoHolic]

...well, nobody can really answer how this will sound.

Regarding oscillations you can have positive or negative effects, depending on the entire setup .... also depending on which other
couplings between the metal and your circuit are existing...
may be connected to ground of your electronics... may be not....


Bye
Markus

 

[Jan Dupont]

This is my personal experiences:
Small capacistances will NOT make a properly designed circuit oscillate, and the benefits of shielding the circuit from outside interference etc. far overrides the negative sides of the small capacistance

 

[destroyer X]

Mr Jan Dupont and also chocoholic

Thanks you all, how are you Chocoholic?

Jan is always precise, short and direct and perfectly clear.

Thank you also Jan.

 

[AKSA]

I agree categorically with Jan.

Cheers,

Hugh

 

[destroyer X]

Yes, i recognize Jan Dupont as one of the best 10 on forum

He is very good, also is clear, when he speak no doubt related what he want to say.

Carlos