[SI-LIST] Re: Noise on BGA core voltage rail
- From: Bob Patel <whizplayer@xxxxxxxxx>
- To: Larry.Smith@xxxxxxx, Anand.Kuriakose@xxxxxxxxxx
- Date: Fri, 9 Aug 2002 10:18:42 -0700 (PDT)
Hi!The best measurement technique I have found for
doing noise analysis on power rails is:
1)Use 50 ohm coaxial cables RG178 with 2.6mm dia.
2)Remove the cap, and solder the cable on the pads.
3)terminate the scope to 50ohm.
Bob
--- Larry Smith <Larry.Smith@xxxxxxx> wrote:
>
> Anand - 1 inch of ground lead on the probe is a lot
> these days. That
> is probably the source of a lot of your noise. You
> want the total loop
> area including your ground lead and hot side of the
> probe to be 1/4
> square inch or less.
>
> I prefer to use 50 Ohm transmission line probes.
> They are much higher
> impedance than any power distribution system that
> you are likely to
> measure. The loop area is measured in square mils.
> Just be sure to
> flip the measurement instrument into the 50 Ohm
> termination state,
> otherwise you will have excessive ringing at the 1/4
> wavelength of your
> transmission line.
>
> One more thing, be sure and remove the decoupling
> capacitor from it's
> pads before you use the pads for measurement. If
> you don't, the
> capacitor mounted on the pads will have a self
> resonant frequency and
> greatly color your noise measurements.
>
> regards,
> Larry Smith
> Sun Microsystems
>
> > From: ANAND KURIAKOSE <Anand.Kuriakose@xxxxxxxxxx>
> > To: pwelling@xxxxxxxxxxxxxx
> > Cc: Larry.Smith@xxxxxxx, si-list@xxxxxxxxxxxxx
> > Subject: RE: [SI-LIST] Re: Noise on BGA core
> voltage rail
> > Date: Thu, 8 Aug 2002 20:19:22 -0700
> > MIME-Version: 1.0
> >
> > Hi philips,
> >
> > Basically i used an FET probe with roughly 1 inch
> ground length b/n
> > probe and test board ground. The measurements were
> taken either directly
> > across the high frequency caps which (quite a few)
> are placed directly under
> > the BGA or on the BGA vias. However i will
> validate the test setup in a
> > manner u have mentioned below.
> >
> > Thanx,
> > Anand.
> >
> >
> >
> > From: pwelling@xxxxxxxxxxxxxx on 08/09/2002 04:16
> AM
> > To: ANAND KURIAKOSE/Apex Data Inc/01@Apex Data
> Inc,
> > Larry.Smith@xxxxxxx@SMTP@Exchange
> > cc: si-list@xxxxxxxxxxxxx@SMTP@Exchange
> >
> > Subject: RE: [SI-LIST] Re: Noise on BGA core
> voltage rail
> >
> > Anand,
> >
> > Have you checked the local probe environment to
> see if you are
> > inductively
> > picking up EMI noise from the components around
> the measurements
> > area? I am
> > already assuming that you have a very short
> ground (less than 1/2
> > inch) from
> > the scope probe ground to a low impedance ground
> on the board under
> > test.
> > Often, we will put ground pads on our boards to
> minimize the loop. I
> > am also
> > assuming that there are not alot of traces run
> between the ground
> > connection
> > and the test measurement point. This is only one
> of the reasons why
> > we limit
> > routing on outside layers of the board.
> >
> >
> > There is an easy test to see if you are creating
> measurement error
> > (the
> > small difference between the 2 measured values -
> may suggest this)
> > by the
> > local environment.
> >
> > That test is to short the oscilloscope probe
> ground to the
> > oscilloscope
> > probe tip. Then move the probe near the area you
> were measuring. You
> > will
> > then "see" the measurement environment as a
> single turn loop. If the
> > noise
> > is there at nearly the same levels, the loop area
> of the ground is
> > suspect -
> > or - look at the rest of the measuring
> environment. The rest of the
> > measuring environment is made up of the probe
> body (in some less
> > expensive
> > probes), probe cable, oscilloscope environment,
> etc... It sounds
> > funny, but
> > if you think of a debug station is , you have
> power supplies with
> > long
> > cables, power cords, Personal Computers, Display
> Monitors, keyboard
> > and
> > mouse cables, etc... I actually walked into a
> situation once with a
> > high
> > resolution monitor at a debug station that
> coupled noise to the
> > oscilloscope
> > probe cable causing a false reading (CRT card
> emissions, and Yoke
> > emissions).
> >
> > Aside from this possibility, you have received
> good advice from the
> > LIST
> > contributors. One thing I wanted to mention is
> that when we think of
> > capacitance from the power plane structure, we
> often consider it
> > high
> > quality capacitance dominated by capaitance not
> inductance. As vias
> > penetrate that plane capacitance structure, the
> planes become more
> > inductive
> > (capillary inductive action) the quality of that
> capacitance
> > degrades. With
> > high pin count BGAs, like the 1152 pin packages,
> really perforate
> > the planes
> > and the buried capacitance suffers.
> Unfortunately, that is where we
> > need the
> > plane capacitance for the higher frequency output
> drivers and core
> > voltages.
> > Creative device fanout helps, and using
> micro-vias help that
> > situation, but
> > are expensive.
> >
> > Has anyone performed research on how the plane
> capacitance is
> > degraded by
> > large BGAs perforations? Both donut style and
> full package
> > perforations?
> >
> > Good Luck.
> >
> > Philip Ross Wellington
> > Mgr. Signal Integrity & EMI
> > L-3 Communications CSW
> >
>
>
------------------------------------------------------------------
> 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 archives are viewable at:
> //www.freelists.org/archives/si-list
> or at our remote archives:
> http://groups.yahoo.com/group/si-list/messages
> Old (prior to June 6, 2001) list archives are
> viewable at:
>
=== message truncated ===
__________________________________________________
Do You Yahoo!?
HotJobs - Search Thousands of New Jobs
http://www.hotjobs.com
------------------------------------------------------------------
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 archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
