> Hi everybody,
> I have a question about the ground and power planes and signal layers that
I must use on my printed circuit board.
> This board  mainly consists of  a virtex 2000 part ( FG1156 ball grid
> I have followed the "xilinx board routability" guidelines and accordingly
the board has six signal layers .
> But I don't know how many ground and power planes I will need ;
> and where in the stack they should be placed with respect to the signal
layers ?

This depends on whether you will be wanting controlled impedance traces,
among lots of other things that the rest of the fellows will be able to
explain better than I...

> The board  uses the following voltages :
> 3.3 Volts : LVTTL I/O's of the FPGA
> 1.8 Volts : FPGA core voltage
> 5.0 Volts : voltage required by other devices like the oscillator etc

This comment is quite concerning to me....
Why are you using a 5.0V oscillator??  If the oscillator is for the FPGA I'd
imagine that the FPGA won't have 5V tolerant clock inputs, and so would
require a 3.3V oscillator.  Best to check before doing the board design...

> Will the presence of different voltages necessitate the use of different
power planes for each voltage ?

That depends on personal preference, how many devices draw power from that
voltage rail and what kind of current they draw, as well as how far apart
they are.
Generally I'd go for the two largest voltages (in terms of current drawn by
devices) and assign a plane to each of them.  Then divide it up for the
smaller (current draw) voltages.  Have the ground planes as being continuous
as much as possible, unless you have mixed signal designs, in which case
divide in two one for digital and one for analog (possibly more depending on
the speed of the digital).

> Moreover, this board involves upto 544 (potentially) simultaneously
switching signals ?

That's alot of signals... If they're all synchronised to the same clock,
this will cause problems (assuming the clock skew is equivalent for all
parts).  If you can possibly alter the clocking of the parts (ie change the
time they switch) then voltage sag would be lessened.  Otherwise you will
need a very good supply decoupling scheme

> Will this affect  the ground and power planes ?
In terms of voltage lost across the ground/power planes then it will indeed.
In terms of EMI then again it will cause a great deal... if you can lag some
of the clocks etc so that the switching is more distributed rather than all
at once this would be reduced.

> I'd really appreciate any suggestions...
> thanks and regards
> Anand Kulkarni

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To leave this list visit:
* http://www.techservinc.com/protelusers/leave.html
* Contact the list manager:
* Forum Guidelines Rules:
* http://www.techservinc.com/protelusers/forumrules.html
* Browse or Search previous postings:
* http://www.mail-archive.com/proteledaforum@techservinc.com
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to