Here's what I've concluded, for now...
Regarding the original question, I think the caps were placed between grounds
to allow isolating a ground if a problem arose. I doubt this is a good design
practice. Another possibility is that in theory, you might want to have a cap
between chassis ground and the ground shield of a very long wire between two
systems, but I haven't found a spec. (for long cables) that accommodates that
kind of shielding so think it's a moot point. I think having capacitors
between grounds is an indication that someone doesn't know what they're doing.
Sorry to disagree, but I'll always have a "GND" symbol in my schematics, even
if I'm designing a satellite. For most signals, it's what I'm going to hook my
ground lead to when measuring voltages. I know the name's not perfect, but it
beats whatever's in 2nd place. For earth system safety issues, I might have
"earth" to handle lightning strikes, and might also have a "GND" (uninsulated,
or green wire) to ensure a return path for chafed cables that short to metal
(to force a circuit breaker to pop).
Thanks for the thoughts,
Jeff Loyer
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On ;
Behalf Of Smith, Larry
Sent: Thursday, November 17, 2016 11:59 AM
To: martin.rowe@xxxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Caps between "isolated grounds"?
I have to go with Bruce on this one. We should eliminate the word "ground"
from our vocabulary unless we are talking about a place for potatoes and
carrots.
Instead, we should discuss the "return path" when we talk about that node in
all of our products with zero volts on it. Then when the "money savers" and
"isolationists" come to signal and power integrity engineers and start talking
about splits we can ask: "What? You are going to cut my return path? How is
the return current going to get between my analog and digital sections? Where
is my PDN return current going to go? How much inductance is that going to add
to my loop?"
Analog designers often make the argument that sensitive analog circuits should
be isolated from digital by cutting the return path between them. But the
analog sections almost always have to communicate with the digital sections
with signal lines. If we cut the on-die return path, possibly the package
return path and heaven forbid the PCB return path, what are those
analog-to-digital signal paths going to do? The A and D reference nodes are
potentially 100's of mV apart and the noise varies with frequency. Is this
what we want? If we are really going to cut the return path between our analog
and digital sections, then we should only use differential signals for all
A-to-D communication.
If we call it what it is - cutting the return path - we begin to comprehend how
much damage we are doing to ourselves with return-path splits.
BTW, if we are talking about safety issues, we should use the word "earth."
Regards,
Larry
PS - These comments apply to signals higher than a few MHz. Sensitive audio
signals where board traces are not long enough to be transmission lines and
long audio cables are a different matter. In these cases, a star topology
makes sense.
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On ;
Behalf Of Martin Rowe
Sent: Tuesday, November 15, 2016 12:39 PM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Caps between "isolated grounds"?
The term "ground" is probably the most misunderstood and misused term in
electrical engineering. I blame the universities. They start their electrical
engineering instruction with DC circuits and then progress to AC circuits with
resistors, inductors, and capacitors. But the ideas of parasitic and
nonschematic effects are seldom discussed in classes. Usually, lab assignments
are relatively low-frequency projects -- probably designed to ensure parasitic
effects aren't encountered.
--Bruce Archambeault
http://www.planetanalog.com/author.asp?section_id=3204
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On ;
Behalf Of Lee Ritchey
Sent: Tuesday, November 15, 2016 3:24 PM
To: dbrooks9@xxxxxxxxxxxxxxx; dmarc-noreply@xxxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Caps between "isolated grounds"?
Sometimes I am prompted to cite Bruce Archambeault from IBM on what ground is.
Here is a quote from him.
"Ground is the place where one plants seeds in the spring time in the hope that
come summer one will get a bumper crop of tomatoes."
The reason for Agnd and Dgnd on an IC is to isolate the analog side of the IC
from the digital side for the purpose of isolating the two circuits
inside the package, not outside it. They should share the same ground
plane on the PCB. Having two different grounds on the PCB does not provide any
performance advantage to the IC.
Outside the IC, efforts should be made to insure the analog source does not
share paths with digital signals. We usually do this by using shielded cables
leading to the analog source.
Hope this clears up some of this confusion.
Lee Ritchey
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On ;
Behalf Of Doug Brooks
Sent: Tuesday, November 15, 2016 8:33 AM
To: dmarc-noreply@xxxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Caps between "isolated grounds"?
Occasionally some people use a capacitor between different grounds in a
misguided attempt to provide to provide a signal return path under a trace that
crosses over the plane split for signal integrity purposes.
This, of course, is usually a bad idea. The purpose for separate "grounds"
in the first place was to isolate noise. All the capacitor does is ensure the
noise has a path between the two grounds.
(Note to Jeff: Please send me a current email address off-line.)
Doug
Loyer, Jeff wrote:
I was looking at some designs and found different ground symbolsby capacitors. Can anyone explain why this might be done? Everything I've
connected
Thanks,
Jeff Loyer
------------------------------------------------------------------
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 forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
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
