calculate the right Cl for crystals? (MCS08AW16)

[Roland]

Hello!

How can I calculate the right capacitance for my µC (MCS08AW16)?

Is it C1 and C2 = 10pF?

I would use the following crystal:

HC-49SM (Hosonic-Electronic):

Shunt capacitance: max. 7pF

load capacitance: min: 10pF

drive level: typ. 100µW => max. 500µW

Rf: 1MOhm

Rs: 0 Ohm

C1: ??

C2: ??

Thank you very much!

Roland

[bigmac]

Hello Roland,

For a parallel resonant crystal with 10pf load capacitance requirement, and assuming equal value capacitors C, the following formula would apply -

(C + Cin)/2 = 10 pF, or

C + Cin = 20 pF

where Cin is the input capacitance to each of the two MCU pins (plus any stray wiring capacitance).

The reason why the crystal manufacturer might specify a minimum load capacitance is unclear.  Normally a crystal will be adjusted during manufacture to the required frequency, for a specific load capacitance value.  Errors in the load capacitance will generally result in relatively small operating  frequency deviations that are probably inconsequential for many applications.

Regards,

Mac

[Roland]

Hello bigmac!

Sorry, I can't find the Cin capacitance from the µC in the datasheet also I don't understand the equations.

Is it possible to use both? Which is the better one?

Should it be the value of C1 and C2 between Cload and 2x Cload?

I tryed also my circuit with two 22pF and 47pF (C1 and C2) - it is working.

Which is the best value for C1 and C2?

Thank you!

Roland

[bigmac]

Hello Roland,

The input capacitance at any I/O pin is specified as 8 pF maximum.  Perhaps increase this to 10pF to make some allowance for stray capacitance.  Using the previous formula, this would mean that each capacitor should optimumly be about 10pF for the crystal you are using.  This value should provide better frequency accuracy.

However, if precise frequency accuracy is not of concern, the capacitor values are not critical, as you have found.  A higher than specified load capacitance will result in a slightly lower operating frequency, perhaps by a kilohertz or two, depending on the frequency of the crystal.

Regards,

Mac