Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Crystals vs. Ceramic Resonators

Status
Not open for further replies.

MacGyverS2000

Electrical
Dec 22, 2003
8,504
0
0
US
I'm bumping up the clock speed on a PIC project (16F72) from the internal 4 MHz to an external 20 MHz. I can use a 20 MHz crystal with two external caps to the tune of $0.80, or a 20 MHz ceramic resonator with internal caps to the tune of $0.40 (which is about half of the size, to boot). Accuracy is not an issue.

However, reading the PIC's datasheet, Microchip lists a max clock speed in the ceramic resonator section at 16 MHz (I'm not sure if this IS the maximum possible, only the maximum listed with tested cap values).

So, other than accuracy over time/temp, is there any advantage/disadvantage to using a ceramic resonator with internal caps compared to a crystal with external caps?
 
Replies continue below

Recommended for you

In my experience with PICs, there is no difference in using both oscillator alternatives. I’m currently driving a design with RC, ceramic and crystal oscillators, at 4MHz (RC and XTAL resonator) and 16 MHz (XTAL + caps).

An important issue is the start up conditions -I mean power supply’s rise time- and using of parallel crystal type like PIC controllers use. Regarding this, in my case, serial crystals do the job as parallel ones in these conditions. Anyway, you should check this.

So, if you don’t have any special requirement, ceramic resonators are OK for most applications.
 
In a medium frequency range, if high accuracy is not a concern, and if your circuit is tuned to it, the ceramic resonator is interesting for cost (at high volumes) and real estate considerations.
 
As a general rule, don't exceed the specified limits of a device. While it may "work" for a while, other unexpected problems may be induced, i.e. increased heat, reduced life span, etc..
 
me,

That's part of the confusion. The device is specified for up to 20 MHz operation. The datasheet lists capacitor values to use as a rough guide for various frequency ranges and oscillator types. The section on crystals lists cap values up to 20 MHz, but the resonator section only lists cap values up to 16 MHz. Now, I'm not saying the datasheet specifies the device will only work up to 16 MHz when using a resonator, only that cap values are specified up to that range.

I was trying to determine if there was something completely different between crystals and ceramics that would lower the possible frequency range. I have found a 20 MHz ceramic with built-in caps that can be had for about $0.30 in bulk. A similar speed crystal with external caps is going to cost me around $0.75 in bulk, and is about twice the size (with caps).
 
Certainly, your issue is mass production. Try also the Microchip Community Forums and manufacturers’ advice regarding parallel oscillators at that frequency range.
 
Those oscillators from Linear are pricey little boogers! 1-100 pricing is $2-$3+. A typical crystal for me, even in single quantities, is just over $1...less than $0.75 in quantity. When you add in their innacuracy, I'm not quite sure what purpose they serve, unless you want to change the frequency on the fly with a resistance change. I suppose the SOT-23 format takes up about half the board space of a typical crystal, though.
 
Status
Not open for further replies.
Back
Top