next topic »

[on:]

hi,
im building something and want to try a resignator and crystal in the same circuit, ive added a pad so i can fit both obviously not at the same time, just wondering if the surface mount pad will effect the resignator/ timing because of the pads?

cheers

[Reply #1 on:]

a resignator

What's that?

if the surface mount pad will effect the resignator/ timing because of the pads?

Yes.

[Reply #2 on:]

Pretty sure he means resonator, probably a ceramic.  You want both just as an experiment?  I've never used those things, they have bad reputation for accuracy and crystals are dirt cheap anyway, why not use a crystal?

[Reply #3 on:]

im building something and want to try a resignator and crystal in the same circuit, ive added a pad so i can fit both obviously not at the same time, just wondering if the surface mount pad will effect the resignator/ timing because of the pads?

In principle, yes (especially if there is a ground plane on the other side of the board); but the effect will be very small, so no need to worry about it.

[Reply #4 on:]

I think there might be more of a problem with the capacitor pads being shorted out by the
metal crystal case, since there's hardly any spacing between the two. 

[Reply #5 on:]

there's hardly any spacing between the two. 

Worse than that, since that doesn't look like a very accurate outline for a typical crystal!
Otherwise, the SMD pads should not affect the accuracy in any meaningful way.  (oh, let's plug in numbers!
Assume a pad 2mm on a side, with a separation of 1.6mm from the other plate, and a dielectric constant of 4 for fiberglass PCB material.)

C = E_r * E₀ * A/d
   = 4 * 9e-12 * (2e-3 * 2e-3)/1.6e-3
   = 0.1 pF (approx)

Here's the layout from one of the official arduino boards, BTW:

[Reply #6 on:]

0.1 pF (approx)

Sounds a lot lower than what I might have expected, but it's a nice formula. I didn't think anything
had stray capacitance that low, :-).

[Reply #7 on:]

Apparently there are also "edge effects" that I didn't take into account:
http://chemandy.com/calculators/rectangular-capacitor-calculator.htm

[Reply #8 on:]

Fringe field effects will increase the value, as will capacitance of the traces going up to the pads, but the general
order of magnitude is correct.  Small things have less capacitance than big things, all else being equal.  Stray
capacitance on interconnect on a VLSI chip is measured in fF (femto farads)!

For comparison a capacitor made from two metal plates sandwiching a credit card would be about 180pF - 0.9mm
thick, assume dielectric constant of 4 (incdentally here the geometry makes fringe-effects negligible).

Most circuit construction techniques mean stray capacitances are a few pF for small parts of the circuit, scaling up
with size.  A ground plane actually reduces the stray capacitance between different signal wires (at the expense
of increasing the capacitance to ground).  For a high-accuracy crystal oscillator you might want to leave a hole
in the groundplane around it to reduce stray capacitances altogether (although for a low-power crystal oscillator
(such as for RTC chip) you want the groundplane to reduce the risk of external interference on the very weak signals.

For a general purpose digital logic clock accuracy isn't of huge concern, so stray capacitance isn't very crucial
here, it will just slightly retune the oscillator.