[SI-LIST] Re: Displaying eye pattern in an oscilloscope?
- From: "Chris Cheng" <Chris.Cheng@xxxxxxxxxxxx>
- To: <si-list@xxxxxxxxxxxxx>
- Date: Sun, 23 Oct 2005 18:03:12 -0700
Let's go off the original topic and talk eye diagram on DDR.
That's an oxymoron by itself. With a single edge strobe, I claim no one knows
when did the data get latch (vref+(noise margin),vref-(noise margin) and
anything in between).
That said, any half decent BIOS software guy can turn off refresh cycle and
write some code to hit 16 address blocks (for 4 bit burst) and 256 address
blocks (for 8 bit burst) in non-interleave mode within a particular DIMM.
If you don't have a BIOS guy, sorry, can't help there.
________________________________
From: Ihsan Erdin [mailto:erdinih@xxxxxxxxx]
Sent: Sun 10/23/2005 8:13 AM
To: steve weir
Cc: Chris Cheng; rrepollo@xxxxxxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: Re: [SI-LIST] Re: Displaying eye pattern in an oscilloscope?
Hi Steve,
I think a simulation vs. measurement comparison is the motivation behind the
eye diagram capturing. I don't know what kind of post-processing filter you can
apply to the captured waveform in order to eliminate one of the 90-degree apart
strobe signals on a bi-directional bus but even if this were possible it would
only work on a single memory interface. In order to understand which memory
device is driving you still need to observe (trigger on) the /CS signal. All
in all, I'm not convinced an eye-diagram scope capture is possible to form a
reference to the simulation results. But if the idea is timing verification as
you put it, yes, you can do that by probing (and logic triggering on) all the
control signals but without an eyediagram pattern on the scope.
Ihsan
On 10/23/05, steve weir <weirsi@xxxxxxxxxx> wrote:
Ishan, I do agree with you that just triggering on DQS and running
infinite persistence will just make an ugly mess on the screen. And if what
one wants to do is detailed characterizations by: target/source DIMM,
interleaved/non-interleaved bursts etc, what I suggest is inadequate.
However, for the very same reason that infinite persistence without a
cycle type trigger doesn't work well, we can extract reads and writes using
only DQS and DQ. The trick is to use the 90 / 0 DQS to DQ phase offset of
write versus read as a post-process filter to separate out the reads from the
writes in .wfm captures. If the idea is to verify that system level timing
works, that should be enough for the data path. If the skew moves around by
more than 90 degrees such that reads can't be distinguished from writes, then
the channel is broken anyway.
Regards,
Steve.
At 11:14 PM 10/22/2005 -0400, Ihsan Erdin wrote:
I'm not sure if it's possible at all. To capture a read or
write cycle, you'll need a logic triggering setup for data and control signals
(/WE, /CS) as Chris describes. For an eye diagram pattern, on the other hand,
you'll need an edge trigger setup either at the clock or the data itself with a
delay (one or two periods) A logic triggering with an infinite persistence
display setting will not result an eye diagram.
Ihsan Erdin
On 10/22/05, Chris Cheng <Chris.Cheng@xxxxxxxxxxxx> wrote:
Not that simple.
You will most like need some advanced triggering to
trigger on CAS, WE so that you can tell if it is a read or write. Then you
probably need some software to limit the hit to certain address to be sure you
get the right DIMM. Good luck if you need to see the effect of interleave vs.
non-interleave.
________________________________
From: si-list-bounce@xxxxxxxxxxxxx on behalf of steve
weir
Sent: Fri 10/21/2005 10:59 PM
To: rrepollo@xxxxxxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Displaying eye pattern in an
oscilloscope?
Don't you have access to both clock and data?
Steve
At 09:17 AM 10/22/2005 +0800, Rep wrote:
>Good day!
>
>I simulated a data signal from SH4 processor to SDRAM,
displaying it in an
>eye pattern.
>I need to verify the signal by actual board
evaluation.
>Could please someone tell me how I could measure the
>signal in eye pattern through the use of an
oscilloscope (TEK)?
>
>I would really appreciated any help I could get.
>
>Thank you.
>
>------------------------------------------------------------------
>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
<mailto:si-list-request@xxxxxxxxxxxxx> with 'help' in the Subject field
>
>List FAQ wiki page is located at:
>
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
>
>List technical documents are available at:
> http://www.si-list.org
>
>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
>
------------------------------------------------------------------
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 FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
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
------------------------------------------------------------------
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 FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
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
------------------------------------------------------------------
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 FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
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:
