measure the xtal

[savnik]

To measure something, anything, you have to have an accurate timebase. To measure the internal clock of the PIC, using that same internal clock as your timebase as you've shown in your brief example, doesn't work.
Do you have an o'scope?

Yes i have.
Or how to measure a xtal on RA4;
I want to measure the frequency of a oscillator with xtal or from a signal connecting to RA4

[skimask]

Yes i have.
Or how to measure a xtal on RA4;
I want to measure the frequency of a xtal or from a signal connecting to RA4

So you want to make a frequency counter then...not measure the internal frequency of your PIC...

[savnik]

So you want to make a frequency counter then...not measure the internal frequency of your PIC...

Yes.I want to measure the frequency of a oscillator with xtal or from a signal.

[skimask]

Yes.I want to measure the frequency of a oscillator with xtal or from a signal.

One of the easiest ways I can think of off the top of my head...

Use PBP's COUNT command...

But that's only good up to about 125Khz when the PIC is running at 20Mhz.

So, you could 'pre-divide' (like the various prescaler/postscalers inside the PIC) the incoming signal using a divide-by-x chip, like maybe a 74193 or something similar.

Feed the input signal into the 74193, take the output from the 74193's Qd output, which will be a divide-by-16 output, and will effectively increase PBP's COUNT commands range up to about 2Mhz (with a decrease in resolution of course).

Cascade 2 of the 74193's back-to-back, and you should be able to measure signals up to about 32Mhz (which is the effective limit of the 74LS193), using PBP's COUNT command.

[Acetronics2]

Hi, Savnik

An easy solution is to use an external RC circuit and use ... RCTime ( 16 Bits result ...) !!!

calibrate RCTime ( low drift components !!! ) with a Known Freq XTal ...

With another XTal The Result will be proportionnal to the XTal Freq change ...

That's all, Folks !!!

Alain

[skimask]

Hi, Savnik
An easy solution is to use an external RC circuit and use ... RCTime ( 16 Bits result ...) !!!
calibrate RCTime ( low drift components !!! ) with a Known Freq XTal ...
With another XTal The Result will be proportionnal to the XTal Freq change ...
That's all, Folks !!!
Alain

Won't the output from RCTime be a logarithmic response, in other highly non-linear? You'd probably have to do some serious math to get accurate results.

[Acetronics2]

Won't the output from RCTime be a logarithmic response, in other highly non-linear? You'd probably have to do some serious math to get accurate results.

Hi, Skimask

you Just forgot The XTal range won't be from zero to infinite ... 1 to 20 ... may be.

But, as always, I'm not so sure of what our friend intends to do ...

Alain