RE: [SI-LIST] : Noise Voltage levels vs. EMI levels

Chris Rokusek (crokusek@viewlogic.com)
Tue, 27 Jul 1999 11:18:15 -0700

David,

> Is there is an absolute noise voltage level that should never be exceeded?

Ultimately there is only the far field requirement. So the trick is to do
what you're trying to do and work back to the differential-mode source (via
the common mode noise voltage).

> Is there a way to "accurately" quantify the impedance's of the PCB ground
> plane and cables?

Two successful methods:

- 3D FDTD (finite difference time domain) can be used to plot
common-mode impedance of the cable configuration over wide frequency
range.

- MEASUREMENT!!

Numerous runs can be made with different cabling configurations (ground
position, shielding, location of connection to board) to come up with
re-usable heuristics.

> Can noise voltage become a PCB design specification?

Better yet the signal that CAUSES the noise voltage can be thresholded in
terms of: signal edge rate, voltage range, loop area, and driver strength
(VI curves), etc.

> Most importantly, when all is said and done, is there an "accurate"
> correlation between noise voltage and EMI levels?

Depends how accurate you need to be and of course how important it is to be
accurate. With the drastic EMI dependence on cable positioning, software,
mfg tolerances, process variation, it may only be necessary (and POSSIBLE!!)
to calculate a worst case number.

Chris Rokusek
Viewlogic Systems

> -----Original Message-----
> From: owner-si-list@silab.eng.sun.com
> [mailto:owner-si-list@silab.eng.sun.com]On Behalf Of Spencer, David H
> Sent: Monday, July 26, 1999 10:04 AM
> To: 'si-list@silab.eng.sun.com'
> Subject: [SI-LIST] : Noise Voltage levels vs. EMI levels
>
>
> I'm looking to correlate PCB ground noise voltage to an EMI levels.
>
> My understanding goes as far as: the voltage is measured from the
> PCB ground
> input to a point on the board differentially. That voltage drop is the
> noise voltage. The voltage drop tells us that there is common mode
> current. If we knew the impedance between the two points we could then
> calculate the common mode current.
> Since every PCB and has different geometry's and cables (power
> and I/O), the
> difficult unknown is the impedance.
>
> My questions:
> Is there is an absolute noise voltage level that should never be exceeded?
> Is there a way to "accurately" quantify the impedance's of the PCB ground
> plane and cables?
> Can noise voltage become a PCB design specification?
> Most importantly, when all is said and done, is there an "accurate"
> correlation between noise voltage and EMI levels?
>
> Thanks for the input
>
> David Spencer
> Xerox Corporation
>
>
> **** To unsubscribe from si-list: send e-mail to
> majordomo@silab.eng.sun.com. In the BODY of message put:
> UNSUBSCRIBE si-list, for more help, put HELP. si-list archives
> are accessible at http://www.qsl.net/wb6tpu/si-list ****
>
>

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****