Thanks for all the very helpful replies. I did some further experiments and
it showed that Aldo's comments hit the root cause of this issue - the
noncausal ripples are due to the insufficient sampling points of the
original sparam. After resampling w/ denser points then rerun the sim, the
ripples disappeared.
One interesting observation is that it seems that the originally
insufficient sampling points cannot be amended by interpolation. I tried
several strategies of interpolation, meanwhile tweaking the fstep sizes,
all to no avail,
Answers to some questions regarding my setup:
1. I used several different ways for causality check:
a. Check the ratio of counter-clockwise phase changes. Thanks to Yuriy's
slides;
b. Check the differences btw the imaginary part of IL and the Hilbert
transform of the real part of IL;
c. Commercial causality check tool.
It seems that a causality OK sparam could actually have very scarce
sampling points.
2. Will the non-causal ripples cause real problems?
My thinking is that the ripples denote that some energies are leaked onto
the time before t=0, which means the remaining energies for the causal part
is reduced. On the other hand, the noncausal ripples will also decrease the
eye margin.
Just some of my thoughts. Any comments will be much appreciated.
Best Regards,
Tramp
On Thu, Jun 11, 2020 at 3:18 PM Morales, Aldo W <awm2@xxxxxxx> wrote:
Hi All:
This is not a problem of the ifft, not fmax. It is a sampling problem as
well as the way your data is being processed. A lot of the papers out there
are wrong. You can contact me to discuss this is in detail. We have
submitted a paper on that regard.
Cheers,
Aldo
_____________________________________________________________________________________________
Aldo W. Morales, Ph.D. IEEE Senior Member
IEEE Consumer Electronics Society 15-16 Distinguished Lecturer
Co-Director, Center of Excellence in Signal Integrity
<http://www.hbg.psu.edu/csi>
https://harrisburg.psu.edu/center-for-signal-integrity
Professor of Electrical Engineering
Penn State Harrisburg
777 West Harrisburg Pike
Middletown, PA 17057
U.S.A.
Phone: (717)-948-6379
Fax: (717)-948-6352
________________________________
From: si-list-bounce@xxxxxxxxxxxxx <si-list-bounce@xxxxxxxxxxxxx> on
behalf of Thomas Beneken <thomas.beneken@xxxxxxxxxxx>
Sent: Thursday, June 11, 2020 10:01 AM
To: si-list@xxxxxxxxxxxxx <si-list@xxxxxxxxxxxxx>
Subject: [SI-LIST] Re: noncausal ripples in impulse response
Hello Tramp,
it's written in the fine print coming with the fft/ifft transformations
that they are valid for repetitive signals, only. The ripple you observe
before t=0 is just a late response to the preceding step pulse. To avoid
that give more time for the responses to the preceding step to settle.
Smaller frequency steps do help.
Another source of ripple is that you just cut off the frequency response
of your DUT at maximum frequency. If the response of your DUT has not
levelled out to zero at maximum frequency there will be ripple in the
time domain. It's like applying a high order low pass filter. The remedy
for this is to have a nice roll off for the frequency response of the DUT.
Thank you!
Thomas
Am 11.06.2020 um 13:43 schrieb Tramp:
Hi experts,doesnt
When obtaining the impulse response using ifft, often the generated ir
(impulse response) shows a bit of ripple&noise before t=0. I did some
experiments and here is the observation:
1. Increasing fmax, which is the max freq for the IL extrapolation,
help reduce the ripple.sparam
2. Changing delta-f shows no avail as well.
3. Although the causality of the sparam is already good enough, still I
tried composing a 'causal' sparam by replacing the imag part of the
w/ the Hilbert transform of the real part. Unfortunately, the thereafterhttps://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.freelists.org%2Fwebpage%2Fsi-list&data
produced ir is a complete mess. What went wrong here?
I read from some literatures where the authors said they simply zero-out
all the ripples before t=0. Is this a recommended approach? Are there
better solutions to obtain a genuine ir?
Best Regards,
Tramp
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
%7C01%7Cawm2%40psu.edu
%7C3b70b66902f3483c6f5908d80e10284c%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C1%7C637274809851554222&sdata=D%2BcQkvpwvVKM%2FxmmiMun552w5lni5oZpBZ6w2FxTXEU%3D&reserved=0
https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Ftech.groups.yahoo.com%2Fgroup%2Fsi-list&data
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
%7C01%7Cawm2%40psu.edu
%7C3b70b66902f3483c6f5908d80e10284c%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C1%7C637274809851564179&sdata=0lPmgoa7fP%2BNim0HcipUDHp5G5altc3IF2TeuyVKY%2Bo%3D&reserved=0
https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.freelists.org%2Farchives%2Fsi-list&data
List archives are viewable at:
%7C01%7Cawm2%40psu.edu
%7C3b70b66902f3483c6f5908d80e10284c%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C1%7C637274809851564179&sdata=yRw3tSXAyONNMqtTf2X7yL9k%2FzD%2B4ICKM%2BETQXa3hHM%3D&reserved=0
https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.qsl.net%2Fwb6tpu&data
Old (prior to June 6, 2001) list archives are viewable at:
%7C01%7Cawm2%40psu.edu
%7C3b70b66902f3483c6f5908d80e10284c%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C1%7C637274809851564179&sdata=MLKNYGj7BPg6Cb8p5pglkn%2F5VD9kw2aN2wF8X%2BLfn%2Fw%3D&reserved=0
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.freelists.org%2Fwebpage%2Fsi-list&data
%7C01%7Cawm2%40psu.edu
%7C3b70b66902f3483c6f5908d80e10284c%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C1%7C637274809851564179&sdata�Yl7HlIhPS5SV6xT2QNwQPMUUiM%2BnVIdOelTgWA%3D&reserved=0
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Ftech.groups.yahoo.com%2Fgroup%2Fsi-list&data
%7C01%7Cawm2%40psu.edu
%7C3b70b66902f3483c6f5908d80e10284c%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C1%7C637274809851564179&sdata=0lPmgoa7fP%2BNim0HcipUDHp5G5altc3IF2TeuyVKY%2Bo%3D&reserved=0
List archives are viewable at:
https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.freelists.org%2Farchives%2Fsi-list&data
%7C01%7Cawm2%40psu.edu
%7C3b70b66902f3483c6f5908d80e10284c%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C1%7C637274809851564179&sdata=yRw3tSXAyONNMqtTf2X7yL9k%2FzD%2B4ICKM%2BETQXa3hHM%3D&reserved=0
Old (prior to June 6, 2001) list archives are viewable at:
https://nam01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.qsl.net%2Fwb6tpu&data
%7C01%7Cawm2%40psu.edu
%7C3b70b66902f3483c6f5908d80e10284c%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C1%7C637274809851564179&sdata=MLKNYGj7BPg6Cb8p5pglkn%2F5VD9kw2aN2wF8X%2BLfn%2Fw%3D&reserved=0
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
