Gary November 29, 2019 at 6:25 pm
- This topic was modified 6 months, 1 week ago by Gary.
I noted that, when entering a span in Spike, I get a different measurement when I actually put cursors on each end. See the attached image. The difference appears to be exactly one bin.
As I understand it, you’re using a set number of points based on the bin width. I have a span of 10 MHz and a RBW of 10 kHz with the Nuttall window. That means I would need a bin width of no more than (10000)/(2.02)=4950 Hz. Given that the bin width must be (80 MHz)/N, where N is a power of 2, the closest match would be N=16384 (2^14). (NOTE: If my calculations are correct, this means that you’re using 16160 points for the RBW, with the remainder of the time-record zero-padded to 16384.) This gives a bin width of (80 MHz)/(16384)=4882.8 Hz. The number of *points* (where a “point” is the center of a frequency bin) is calculated as span/(bin width), which in this case is (10 MHz)/(4882.8 Hz)= 2048. However, because of the fact that the span is essentially zero-indexed (the first point is essentially a bin value of 0), you wind up losing a half of a bin (only the upper half of that first bin on the left counts), while the last bin (all the way on the right) also loses the upper half of the bin (only the lower half counts). All told, you wind up losing a full bin.
I’d like to make two suggestions:
1) Add one more to the point count when calculating the number of points to display the span. In the case above, the point count should be 2049, not 2048.
2) Show the actual span, not the user-entered one, so that there’s no misunderstanding of the span being shown.
Thanks, and I do love Spike and my BB60C!
Attachments:You must be logged in to view attached files.
AndrewModeratorAndrew December 2, 2019 at 6:39 am
Thank you for the thoughtful suggestions. I can’t guarantee we will make any changes based on your suggestions, but it sure is fun to read and think about. I think there are a lot of valid ways to approach the problem and I’m sure other companies take a slightly different approach, such as plot an additional point.
Thanks again for taking the time to provide us with feedback and suggestions!
You must be logged in to reply to this topic.