[SI-LIST] Re: Shield ground isolation - logic/chassis ground connection

All,

For circuit packs and chassis, should the pack's logic ground planes be
contiguous and connected to the metal faceplate with plated through-holes
thus connecting logic ground to chassis ground?  Currently, the packs have a
chassis ground planes along the front edge of  the pack which provide holes
for mounting the bracket.  Logic grounds are separated from this chassis
ground with a gap.  I've placed pads at each of the five through-holes for
the bracket to allow logic and chassis to be connected or not based upon EMC
testing.  

There are some unshielded I/O which are filtered using transformers and
chokes.  Connectors are metal and connected to the faceplate (chassis) by
contact.  I've had good luck using this scheme during EMC testing.

The idea of having the option to isolate chassis from logic ground is for
ESD, EFT, etc. immunity.  Those test signals are referenced to chassis
ground and keeping all offending currents on the outside of the shield
(enclosure) would boost the system's immunity.

Thanks in advance for any assistance.  Grounding gets me every time.  What
to do with logic grounds and chassis for systems that operate into several
hundred MHz and has unshielded I/O ports.

Jack

-----Original Message-----
From: npischl <npischl@xxxxxxxxx> [mailto:npischl@xxxxxxxxx]
Sent: Monday, February 24, 2003 2:05 PM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Shield ground isolation



Annand,

In addition to other good replies you received on the list:

With shielded cabling, for low emission you need to minimize common-
mode (CM) currents that flow on the exterior of the shield. In 
general, is not true that it is achieved by disconnecting the logic 
ground from shield ground. What is needed is high transfer impedance 
of the overall shield. Transfer impedance of a shielded cable for 
example is the (intentional signal) voltage between the center 
conductor and the shield divided by the CM current that flows on the 
exterior due to imperfections. If the cable shield is perfect, and 
the source and the load are perfectly terminated with 360-degree 
connection around the ends, the transfer impedance is very high 
(no "leakage") and the emission is low.

You can easily expand the concept to the overall shield, including 
the "box", except that the definition of the driving voltage may 
become little more ambiguous. In most systems you will deal with, the 
two main causes of increased emission from unshielded cables will be 
imperfect connection of the shield at the connectors (pigtails), or 
compromised shielding-integrity of the chassis (e.g. a slot). They 
both let CM currents flow on the exterior of the shield and cause 
emission.

Trying to "disconnect" the two metals (digital and shield) will not 
work well because, you want it or not, they are connected by their 
mutual impedance (C in the first place) anyways. This impedance lets 
current flow between them. This may also create serious functional 
problems when the cable shield is used as a reference (I have seen it 
unfortunately many times).

It is important to keep continuous and low-impedance current path for 
both the intentional and CM signals, while minimizing external CM 
currents (maximize transfer impedance). This can be done by 
connecting the signal GND and the shield together at the connector, 
and maintaining low-impedance continuous connection for the 
intentional signals and 360 degree (or best approximation) of the 
cable-shield. The result is that the intentional signal current 
returns through low-impedance and that CM voltage at the connector 
(mainly the voltage between the PCB GND and the shield) is low, thus 
CM current on the cable is also minimized. What it does is that it 
lets CM current return to its source withini the box, instead of 
getting on the exterior of the shield.

If doing so doesn't reduce emission from the shielded cable, you need 
to look at the shield-integrity of the box that is most likely 
compromised. Since we live in an imperfect world, sometimes you may 
see it happen, but it does not prove that disconnecting the signal 
from the shield is necessary for good EMC. Quite opposite is 
generally true.

The only exception that may sometimes validate disconnecting the two 
may be if you drive the I/O through transformers or some other means 
of isolation. Even then, it depends on how much isolation (e.g. C 
between primary and secondary or across an optocoupler) you really 
have.

Neven





--- In si-list@xxxxxxxxxxxxxxx, "Kuriakose, Anand" 
<Anand.Kuriakose@xxxx> wrote:
> 
> Hi All,
> 
> On high speed boards with I/O connections running to outside world, 
the
> shield ground of the connectors should be isolated from the digital 
ground
> if the board has to meet the requirements of EMI/EMC compliance 
standards.
> 
> Are there any side-effects of isolating the shield ground from the 
digital
> ground?
> 
> 
> I have seen on some boards, that the isolation is achieved by using 
ferrite
> beads. On few others i have seen capacitors being used b/n shield 
and
> digital ground.
> 
> How are capacitors useful in acheiving isolation? Or are these caps 
used for
> some other purpose, which i am not able to understand? 
> 
> What is the best method to achieve this isolation?
> 
> Thanks in advance.
> 
> Regards,
> Anand.
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxx with 'unsubscribe' in the Subject field
> 
> or to administer your membership from a web page, go to:
> http://www.freelists.org/webpage/si-list
> 
> For help:
> si-list-request@xxxx with 'help' in the Subject field
> 
> List archives are viewable at:     
>               http://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

------------------------------------------------------------------
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:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:     
                http://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
  
------------------------------------------------------------------
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:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:     
                http://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: