About Analog Discovery 3_Pattern Generator_clock accuracy

[ryoryo10]

If the Pattern Generator outputs a 40kHz clock, what is the frequency accuracy or error?

[JColvin]

Hi @ryoryo10,

I might be mistaken, but I don't believe Digilent has defined this particular parameter.

I likely did not do a proper setup or test, but with the AD3 at my desk outputting a 40 kHz clock from one of its channels, this is what I measure on the ADP2230 oscilloscope (using a different instrument purely to show a different internal oscillator being used for timing) after over 5000 acquisitions at 100 MHz sample rate (with the default 32768 sample per acquisition size, this works out to about 65k 40 kHz pulses being analyzed):

When the precision is scaled to the 1 mHz level (the first row in measurements), I get (to be clear, with a single AD3 on a singular hardware setup) an average measured frequency of 39,999.669 Hz, or -8.275 * 10^-4 % less than a precise 40 kHz.

When I ran a similar setup (again, one unofficial, probably poorly setup singular test on a singular hardware setup) for a 1 MHz digital square wave, the difference between the measured average and the theoretical was very similar at -8.273 * 10^-4 % less, though the measured minimum and maximum were less close when compared to the 40 kHz counterpart.

Let me know if you have any questions.

Hope this helps,
JColvin