From Friday, April 19th (11:00 PM CDT) through Saturday, April 20th (2:00 PM CDT), 2024, ni.com will undergo system upgrades that may result in temporary service interruption.

We appreciate your patience as we improve our online experience.

LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Theoretical question on downsampling

Hey guys,

 

I have a samplerate of 48kSPS on the myRIO. After a phase selective detection, the wanted signal is at 0Hz = DC. Now, I have to downsample my signal, as I want to low pass filter the signal at around 1Hz or 0.5Hz.

For this I use a 2 stage CIC filter. The input is I24.24 converted into I64 and as I only use change rates which are powers of two, I scale my CIC output with a bitshift back to I24.24, so register growth is no issue for further processing.

 

Can someone tell me what are disadvantages of high decimation rates?

 

At the moment I use 512, but I cant find any disadvantage in using 1024 or 2048...

 

Kind regards

 

Slev1n

0 Kudos
Message 1 of 5
(2,932 Views)

Wow!  This went right over my head!  I looked up CIC filters (found a little, at least I know now what they are, but not sure when to use them), but am still puzzling out I24.I24 (fixed-point format?  unsure ...).

 

Slev1n -- would you mind educating me a bit on these topics (or providing a helpful link)?

 

Bob Schor

0 Kudos
Message 2 of 5
(2,902 Views)

It would be nice if there was even a mention of "LabVIEW" somewhere in the post.  What does this have to do with LabVIEW?

Bill
CLD
(Mid-Level minion.)
My support system ensures that I don't look totally incompetent.
Proud to say that I've progressed beyond knowing just enough to be dangerous. I now know enough to know that I have no clue about anything at all.
Humble author of the CLAD Nugget.
0 Kudos
Message 3 of 5
(2,872 Views)

This has not much to do with LabVIEW, but I thought maybe someone knows this here as digital signal processing is often used in LabVIEW, too.

 

here is a link to a paper from the "inventor" of CIC Hogenauer: CIC

 

IXX.YY: XX stands for the bit width of the figure, YY how many bits are used to represent the integer part. XX-YY, how many bits are used fpr fractional representation.

0 Kudos
Message 4 of 5
(2,846 Views)

Fair enough.  Clearly, among the people who have already responded, you probably know far more about this topic than we do.  So here's a suggestion -- why not play "scientist" instead of "engineer"?  Do an Experiment -- take a known signal (real or simulated), add some simulated "noise", and try out various decimation rates to see if you notice any difference.  You might even push the limits to see where it breaks down, and if you can relate that to some "parameter of interest".  If you learn something useful, you can respond to your own Post, marking your Reply as the Solution (in case someone else with a similar question does a search on "Downsampling" and finds your post, now with a Solution).

 

Bob Schor

Message 5 of 5
(2,818 Views)