Re: FW: [SI-LIST] : Comments from your SI seminar (SendII)`

About this list Date view Thread view Subject view Author view

From: Ronald Miller (rbmiller@sjm.infi.net)
Date: Mon Nov 01 1999 - 10:37:47 PST


> Mr Grasso
>
> Ground currents flow in adjacent layers to the signals they mirror. Back in
> the old days it was found that by seperating grounds it was possible to keep
> these ground currents seperated from each other and thereby reduce the
> interference
> between power supply and other noise cources from signal circuits.
>
> Well, that was great for frequencys up to a few Mhz and for simple systems.
> However,
> about 25 years ago in microwave radio we found that it became more and more
> difficult to isolate because of the many signals that had to go from one chassis
> or module
> to another, like video on coax and control signals which needed a good ground
> return.
> Differentialshileded pairs were used as much as possible and the shield was
> floated
> at one end to minmize inductive coupling from the ground to the pair internal.
>
> However, as frequencys increased also, it became apparent that this could
> no longer work. There was capacitive coupling everywhere, and this was
> impossible
> to eliminate.
>
> So, since all the grounds are connected anyway through capacitance and mutual
> inductance the industry moved towoard minimizing the interference caused by the
> coupling. We tied all the grounds together with as much copper as possible in
> order
> to reduce the resistance and inductance in these ground paths, thereby reducing
> the
> voltages across these planes and thereby reducing the voltages induced in the
> ground
> paths of the more sensitive circuits. Tieing the chassis to the PCBs reduces
> this
> coupling further, with as many grounds as possible. Several grounding bars
> across
> the bottom of a pcb tied to the chassis continuously is nearly optimal.
>
> For high power microwave amplifiers the ground layer often becomes 1/4 inch
> aluminum
> plate. So doing really stabilizes things and control circuits are easily kept
> from the
> effect of the adjacent RF.
>
> Recently, I applied this technique to a product to reduce EMI, and it was a hard
>
> battle to get the people convinced, but the final product was many db better
> than
> the previous product that no one questions it.
>
> However, when I lay out power circuits on the same board with signal circuits,
> I use a certain ammount of moating of the ground to steer the Power currents
> away from the
> sensitive circuits. The grounds are still connected but the ground currents are
> kept
> isolated as much as possible.
>
> When I have a ground plane on top and bottom of a pcb, and stitch it together
> around the edges I have created a shielded enclosure and the rf on the internal
> traces cannot escape so is absorbed in the dielectric material. The spacing as
> you mentioned should be about 1/20 of a wavelength at the highest frequency
> of intrest.
>
> So, welcome to the new methods Chaz.
>
> Ron Miller
>
> "Grasso, Charles (Chaz)" wrote:
>
> > Really?? Then why not discuss these issues in a public forum?
> > It seems a consortium of opinions is forming with no discussion
> > allowed. Seems like dogma to me..
> >
> > Thank you
> > Charles Grasso
> > StorageTek
> > 2270 Sth 88th Street
> > Louisville CO 80027
> > Tel: (303)673-2908
> > Fax(303)661-7115
> >
> > > ----------
> > > From: Ron Miller[SMTP:rmiller@Brocade.COM]
> > > Sent: Wednesday, October 27, 1999 1:51 PM
> > > To: si-list@silab.eng.sun.com
> > > Subject: Re: [SI-LIST] : Comments from your SI seminar (SendII)`
> > >
> > > I hope everyone goes to the seminar, because I will be the only person
> > > left
> > >
> > > in this field that understands what is really happening.
> > >
> > > Ron Miller
> > >
> > >
> > >
> > > In regards to comment 1. In a high frequency signalling environment, The
> > > return current will flow on the nearest plane whether it is power or
> > > ground.
> > > If you have two planes coupled closely togther you create a capacitor
> > > which
> > > the high frequency return current uses to traverse the planes and takes
> > > the
> > > path of least impedance.
> > > If there were not two planes coupled together, it would be an EMI
> > > catastrophe and you would not know where the return current is flowing.
> > >
> > > Regarding comment 2, he was stating that grounding in multiple locations
> > > is
> > > a bad idea. This can create some ground loops within the chassis. There
> > > will be some type of potential difference from one chassis connection
> > > point
> > > to another and this will create some current in the chassis. This is very
> > >
> > > bad from an EMI perspective.
> > > He believes one Solid connection to the ground plane is sufficient.
> > >
> > > I
> > >
> > > --
> > > Ronald B. Miller _\\|//_ Signal Integrity Engineer
> > > (408)487-8017 (' 0-0 ') fax(408)487-8017
> > > ==========0000-(_)0000===========
> > > Brocade Communications Systems, 1901 Guadalupe Parkway, San Jose, CA
> > > 95131
> > > rmiller@brocade.com, rbmiller@sjm.infi.net
> > >
> > >

**** 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 ****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Tue Feb 29 2000 - 11:39:28 PST