sample rate math. need proof of quality.

richhutchins · Post by **richhutchins** » Thu Jan 30, 2003 2:36 pm

I get this question alot when I talk to guys at work.

"Hey that displayed wave from your ADC212-100 looks great on your laptop, but I wonder how many sample points are on the curve? If your sample rate is too slow, you might miss a narrow spike or alias the signal."

So, I know that it is capable of 100 Mega samples per sec, but this must vary some how in accordance to the on-board memory. I would like a math formula that, based on the settings in PicoScope, would tell me how many horizontal samples are taken in a given domain of time.

I think it is simply: (100x10e6)/duration of time in full window. But, at what point does the memory become a limiting factor?

Also, please check my concept of vertical resolution. If my voltage scale is 10v Pk-PK, then the smallest detectable change in the signal would be:

10v/(2e12)= .00244......volts or 2.441milliVolts.

matthew · Post by **matthew** » Fri Jan 31, 2003 12:44 pm

Hi Rich,

You can set the maximum number of samples per scope trace in the Settings|Options|Advanced dialog. This will give you an idea of the time-resolution. Alternatively you can save the trace as a .txt file and count the number of readings in a spreadsheet.

Your concept of vertical resolution is correct, except that the notation would be (for a ±5V range):

Code: Select all

10V / 2^12

2e12 implies 2 x 10^12 which is different.

Kind Regards,