[SI-LIST] Re: [Possible SPAM] EMI coupling of clock on other signals

From: "Marc Humphreys" <mhumphreys@xxxxxxxxxxxxxxxxxxx>
To: "chundi srikanth" <chundis@xxxxxxxxx>, <si-list@xxxxxxxxxxxxx>
Date: Fri, 22 Jan 2010 10:30:02 -0800
Sree,
I'd start with trying to pin point the source - before stabbing at 
remedies. Assume for now that the source is either the clock chip and or 
oscillator thats driving it.
Once the source is located then you can start working on identifying the 
couple mechanisms between the source and the rest of your brd

A few suggestions on doing that.
First, what do you have for power filtering on your oscillator? You may try 
isolate it from the power rail with a small inductor, if you have none on 
hand, even the inductance from a zero ohm 0402 surface mount resistor can 
be effective. 
Second, try tame your clock edges by adding or increasing series resistors  
on all clock 83.33MHz outputs from the oscillator and clock chip - any 
standard values between 22 and 47 ohms would be a good starting point.
Third, you could disconnect the output from the clock chip that drives the 
SDRAM and scan the brd again to help determine if its the SDRAM or the 
clock sources that are responsible for the problem. 
Fourth, also you scan the brd with no activity to the SDRAM see if the 
spectrum changes at all - kill the CPU clk if neccessary?

If you could share a little more info on  your brd stackup and power 
distribution planes, use of split planes etc. that info may be helpful down 
the line. 

Marc

-------------------------------------
 Marc Humphreys
LogicLink Design, Inc
-------------------------------------

----------------------------------------
 From: "chundi srikanth" <chundis@xxxxxxxxx>
Sent: Thursday, January 21, 2010 1:25 PM
To: si-list@xxxxxxxxxxxxx
Subject: [Possible SPAM] [SI-LIST] EMI coupling of clock on other signals

Hi folks,
Iam working on a board which is of 16-layer and its a mixed signal board
which is having signals running at around 3GHz. Iam testing this board 
with
initial software. And we have one clock chip whose input is 30.72MHz and 
its
outputs are at 122.88MHz. And This board got an MCU whose highest running
frequency is 166MHz. And it was interfaced to a 32-bit SDRAM and it was
running on a clock speed of 83.33MHz. Today when i probed the 122.88MHz
clock out in a spectrum analyzer there is a strong spur of frequecy
components which were at 166.7, 215.8, 332.4MHz, 489.7MHz etc..... These 
are
almost harmonics of SDRAM clock frequency. And everywhere even if i probed
the GND the components can be seen. So what might be the problem and can 
you
suggest what can be done on board to resolve this.

Thanks in advance for your help.

Regards
Sree

------------------------------------------------------------------
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 technical documents are available at:
http://www.si-list.net

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

 


------------------------------------------------------------------
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 technical documents are available at:
                http://www.si-list.net

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
[SI-LIST] Re: [Possible SPAM] EMI coupling of clock on other signals

Other related posts:
