[SI-LIST] Re: Flight Time

  • From: Bob Patel <whizplayer@xxxxxxxxx>
  • To: brian.p.moran@xxxxxxxxx, si_list <si-list@xxxxxxxxxxxxx>
  • Date: Wed, 13 Nov 2002 18:31:20 -0800 (PST)

Hi! Brian, In my case the clock and data are being
driven from the same source since it is demuxed and
drive the same load.
The problem I have is that I can run the interface at
two different speeds and the clock-data delay times,
setup/hold times are different for both speeds.
Anyway, based on calculation with all the skew & guard
band added the Tflight min for one speed is 1.28ns &
the other speed it is 0.73ns. Hence the minimum length
for the traces come out to be 7.23" and 4.124". The
obvious answer would be to keep the length at 7.5" but
then I violate the max len for the higher speed which
is 6.2".
So you see I am kinda stuck. Any suggestions.
Thanks
Bob
--- "Moran, Brian P" <brian.p.moran@xxxxxxxxx> wrote:
> 
> Hi Bob,
> 
> Its not always the case that signal and clock should
> be equal length for
> best margins, but its a reasonable place to start.
> This is the case when
> clock and data or strobe and data are identical in
> topology and loading,
> assuming the data valid to strobe and data valid
> after strobe specs are
> fairly symetrical.  In other cases you really have
> to solve the timing
> margins for setup and hold using all four flight
> corners, clk-min, clk-max,
> data-min, data-max, and then equalize your setup and
> hold margins. The
> differences in loading and topology can create
> offsets between optimal clock
> length and optimal data length. You are correct of
> course that any positive
> margin can be traded off for length mismatch once
> centered.  
> 
> Also, I don't always use skew as a metric. Sometimes
> I use skew and
> sometimes I use flight time, depending on how the
> margin formulas are
> written, and the characteristics of the two signals
> in question.  We tend to
> use skew based formulas for closely matched source
> synchronous busses where
> data and strobe have identical buffers and toplogies
> and your just looking
> for interconnect and buffer induced skew. In this
> case skew is a tightly
> controlled metric which correaltes to bus
> performance limits. In other cases
> such as clock to address or clock to control where
> topologies and loading
> differences create very large differences in flights
> we just go with flight
> time based analysis, since skew isn't a meaningful
> metric.    
> 
> Brian P. Moran
> Signal Integrity Engineer
> Intel Corporation
> brian.p.moran@xxxxxxxxx
> 
> 
> -----Original Message-----
> From: Bob Patel [mailto:whizplayer@xxxxxxxxx]
> Sent: Wednesday, November 13, 2002 2:57 PM
> To: si_list
> Subject: [SI-LIST] Flight Time
> 
> 
> 
> Hi! I had a question regarding Flight time. If I
> calculate the maximum flight time to meet the setup
> time and minimum flight time to meet hold time ans
> say
> they are maxFlt time=4ns and min Flt time=1.28ns.
> This is for a source synchronous bus, then what
> should
> I keep the trace length for clock & parallel 15 bit
> data bus.
> I am saying I can keep the clock bus around the same
> length and the length between the min & max flight
> times.
> Do you'l see any problems?
> THanks
> Bob
> 
> __________________________________________________
> Do you Yahoo!?
> U2 on LAUNCH - Exclusive greatest hits videos
> http://launch.yahoo.com/u2
>
------------------------------------------------------------------
> 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
>   
>
------------------------------------------------------------------
> 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
>   
> 


__________________________________________________
Do you Yahoo!?
U2 on LAUNCH - Exclusive greatest hits videos
http://launch.yahoo.com/u2
------------------------------------------------------------------
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: