Re: [PEDA] Revision problems

2002-08-26 Thread Michael Reagan (EDSI)

Dennis
The reason you want to use update afterwards is because you want to update
relative to the  latest netlist.

Mike

- Original Message -
From: Dennis Saputelli [EMAIL PROTECTED]
To: Protel EDA Forum [EMAIL PROTECTED]
Sent: Sunday, August 25, 2002 11:52 AM
Subject: Re: [PEDA] Revision problems


 shouldn't you load the netlist AFTER the update?
 else stray tracks could pick up connectivity?
 i haven't found clearing nets necessary, although i do it sometime

 Dennis Saputelli


 Michael Reagan (EDSI) wrote:
 
  Robert,
 
  Update free primitives is a very good command to use.  I use flawlessly
on
  every design.  If you are getting drc errors when invoking the command
then
  you must have traces shorted from one  pad to another.
 
  The exact process I use is different than most others recommended
practice,
  here is what I do
 
  1 clear all nets from the design
  2. load a new netlist
  3 update free primitives
  4 run drcs
 
  The reason I follow this procedure is because it can save hours of
Protel
  crunching the netlist over and over again.   If the pcb still has drc
errors
  then there is a problem in the pcb file.
 
  I have used this process for hundreds of designs, and hundreds of
thousands
  connections.   It has never failed me
 
  Mike Reagan
  EDSI
  Frederick MD
 
  - Original Message -
  From: Robert M. Wolfe [EMAIL PROTECTED]
  To: Protel EDA Forum [EMAIL PROTECTED]
  Sent: Sunday, August 25, 2002 9:04 AM
  Subject: Re: [PEDA] Revision problems
 
   To All on this one,
   I have also been in this situation
   with a board I had to use the update free primitive
   to connected copper. IMHO this is a VERY VERY dangerous command unless
you
   are ABSOLUTELY sure there is nothing touching any pads that it should
   not be. I have used fottprints of pads touching to
   tie A  D GND's etc together and if that has been done
   it will pick one of those nets and make that one net.
   NOT just the fact that trace touched this was clearly
   only pads touching but it made both ends one net.
   Also if there were footprint changes and traces
   just happen to be touching things that really don't
   want to be you are also in trouble. Yes DRC should
   flag all of these but IMHO this is a very dangerous
   and time consuming way to deal with this.
   I had to use this function because I did a board
   with contact patterns for switches I built into
   the footprint, and this is the only way Protel deals
   with making those tracks in the footprint become
   proper net. Again don't like it at all this way.
   It also seems to me that Protel is not
   consistant with respect to update, I have seen
   in some cases what is described in this thread,
   and also times where it actually updated the nets
   properly to new connections. Also it does not seem
   consistant with update footprints either, sometimes
   it does it and sometimes I have had to update
   them manually as I knew I changed one and it was not updated during
PCB
   update with that option checked.
   Also on this board I did see that Online DRC did
   not actually update in real time, only after moving aound
   the board, and if I remember correctly, and maybe closing and going
back
  in
   did I in fact see the green
   come up to inform me I had the shorts.
   I have not dealt with that phone board in awhile
   and my changes to boards has been pretty
   minimal so I have not seen this in awhile, but
   I would say Protels PCB update is a little
   flaky, and a word of caution be very careful
   with the update free primitives command
   Bob Wolfe
  
  
   - Original Message -
   From: Dennis Saputelli [EMAIL PROTECTED]
   To: Protel EDA Forum [EMAIL PROTECTED]
   Sent: Friday, August 23, 2002 9:25 PM
   Subject: Re: [PEDA] Revision problems
  
  
i am on the same page as you about this
and yes almost all of our bds have polygon pours, so that is
probably at
the root of the 'false DRCs'
   
i always reload the netlist after a copper update, just a sanity
check
   
and now for a confession
i STILL don't use the synchronizer
i have been told i am missing something there and i don't doubt it
   
i just never seem to have the time to learn anything and a lot of
our
netlists are from orcad with all the wrong footprints
   
Dennis Saputelli
   
Brad Velander wrote:

 I get those not really a DRC problem highlighing all the time.
  Drives
   me
 nuts. I can change the size of a pad that is connected to a pour
   (legally)
 and the whole pour lights up as a DRC. I usually repour the
polygon
   because
 it is faster than running the DRC and the highlight goes away.
When I
   see
 it, it is usually on a polygon pour.

 Yeah, I would expect that your check would work with the
   reloading
 of the netlist as well. I don't think that you mentioned the
reload of
   the
 netlist in the first reply so he might have had a real 

Re: [PEDA] Revision problems

2002-08-26 Thread Robert M. Wolfe

Thanks Guys,
I see all of your points, I agree with you that
once you understand how it works
you can use it safely. (it can be a time issue if
you get out of wack)
I was trying to point out that if you are not
careful and like you said understand what
it will do it can take a bit of time to get things
back on track. A lesson I learned the hard
way on that early design when I was struggling.
I have also used it since but was surprised
recently (like I said earlier)
with the fact that it looked at pads
touching and changed nets, but now that I know
what to look for I can prepare better. It was only in one place on that
design so it was quick to fix.
It really comes down to the extra steps (time) to deal with these things,
other systems I've used did deal with this a bit more gracefully.
Sorry if a scared the original poster on this one.
Bob Wolfe

- Original Message -
From: Dennis Saputelli [EMAIL PROTECTED]
To: Protel EDA Forum [EMAIL PROTECTED]
Sent: Sunday, August 25, 2002 2:52 PM
Subject: Re: [PEDA] Revision problems


 shouldn't you load the netlist AFTER the update?
 else stray tracks could pick up connectivity?
 i haven't found clearing nets necessary, although i do it sometime

 Dennis Saputelli


 Michael Reagan (EDSI) wrote:
 
  Robert,
 
  Update free primitives is a very good command to use.  I use flawlessly
on
  every design.  If you are getting drc errors when invoking the command
then
  you must have traces shorted from one  pad to another.
 
  The exact process I use is different than most others recommended
practice,
  here is what I do
 
  1 clear all nets from the design
  2. load a new netlist
  3 update free primitives
  4 run drcs
 
  The reason I follow this procedure is because it can save hours of
Protel
  crunching the netlist over and over again.   If the pcb still has drc
errors
  then there is a problem in the pcb file.
 
  I have used this process for hundreds of designs, and hundreds of
thousands
  connections.   It has never failed me
 
  Mike Reagan
  EDSI
  Frederick MD
 
  - Original Message -
  From: Robert M. Wolfe [EMAIL PROTECTED]
  To: Protel EDA Forum [EMAIL PROTECTED]
  Sent: Sunday, August 25, 2002 9:04 AM
  Subject: Re: [PEDA] Revision problems
 
   To All on this one,
   I have also been in this situation
   with a board I had to use the update free primitive
   to connected copper. IMHO this is a VERY VERY dangerous command unless
you
   are ABSOLUTELY sure there is nothing touching any pads that it should
   not be. I have used fottprints of pads touching to
   tie A  D GND's etc together and if that has been done
   it will pick one of those nets and make that one net.
   NOT just the fact that trace touched this was clearly
   only pads touching but it made both ends one net.
   Also if there were footprint changes and traces
   just happen to be touching things that really don't
   want to be you are also in trouble. Yes DRC should
   flag all of these but IMHO this is a very dangerous
   and time consuming way to deal with this.
   I had to use this function because I did a board
   with contact patterns for switches I built into
   the footprint, and this is the only way Protel deals
   with making those tracks in the footprint become
   proper net. Again don't like it at all this way.
   It also seems to me that Protel is not
   consistant with respect to update, I have seen
   in some cases what is described in this thread,
   and also times where it actually updated the nets
   properly to new connections. Also it does not seem
   consistant with update footprints either, sometimes
   it does it and sometimes I have had to update
   them manually as I knew I changed one and it was not updated during
PCB
   update with that option checked.
   Also on this board I did see that Online DRC did
   not actually update in real time, only after moving aound
   the board, and if I remember correctly, and maybe closing and going
back
  in
   did I in fact see the green
   come up to inform me I had the shorts.
   I have not dealt with that phone board in awhile
   and my changes to boards has been pretty
   minimal so I have not seen this in awhile, but
   I would say Protels PCB update is a little
   flaky, and a word of caution be very careful
   with the update free primitives command
   Bob Wolfe
  
  
   - Original Message -
   From: Dennis Saputelli [EMAIL PROTECTED]
   To: Protel EDA Forum [EMAIL PROTECTED]
   Sent: Friday, August 23, 2002 9:25 PM
   Subject: Re: [PEDA] Revision problems
  
  
i am on the same page as you about this
and yes almost all of our bds have polygon pours, so that is
probably at
the root of the 'false DRCs'
   
i always reload the netlist after a copper update, just a sanity
check
   
and now for a confession
i STILL don't use the synchronizer
i have been told i am missing something there and i don't doubt it
   
i just never seem to have the time to learn anything and 

[PEDA] Tearing my hair out with DXP

2002-08-26 Thread Tim Hutcheson

All,

For those that might be trying the multichannel feature of DXP:  I just
learned the hard way (hours of fiddling) about a few confusing things with
the method.  The little bit of documentation in the manual, Introducing
Protel DXP, creates a couple of misunderstandings that I could only resolve
by taking the 8-channel Mixer example apart.  So here are some lasting
impressions from the experience, in the order of points presented in the
manual.

First, when you place a Sheet Symbol and add the Repeat(CIN,1,8) style
Designator, DXP DOES NOT create a Sheet Symbol that looks like the one in
the Mixer example with multiple overlapping sheets.  These appear to be just
artwork added to the mixer schematic as rectangles for dramatic effect and
the example would better without them.  I know it is a minor point, but when
other things don't work as they probably will not, it left me thinking that
I hadn't declared something correctly to invoke the full power of the
repeat.  True, it is shown correctly in the manual diagram on page 6-2 but I
was hard to convince (for about an hour or so, much later, when trying to
get all to work by generating a netlist without error).

Second, the technique of representing the bus coming out of a Sheet Entry
symbol that has a Repeat as a single wire which then connects directly to a
bus, (so that individual signals can be picked out), is so crucial but not
obvious looking at the example diagram on page 6-2.  There is only a small
stub of bus wire connected to the Headphone wire in the picture and I
overlooked that for quite a while.

Third, it is possible to have the designators in the sheets incremented
sequentially as _1, _2, ..._n, by simply leaving out the first parameter of
the Designator Repeat() statement.  I think this will mean that the
technique of repeating channels on the circuit layout (assuming you route by
hand) by simply copying the first one multiple times should still produce
the same orderly incrementation of the designator in each channel.  But I
haven't verified this yet.

.end - of - rant.

regards,

Tim Hutcheson
Research Associate
Institute for Human and Machine Cognition
University of West Florida
40 S. Alcaniz St.
Pensacola, FL 32501
USA
805-202-4461



* Tracking #: 471375E444F874419B41CCFBFC51A3346C4D8132
*


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



[PEDA] Protel 99SE license

2002-08-26 Thread Lloyd Good


Hello all,
I have a client that is looking for a Protel 99SE license. Does anyone have
one for sale. Please email me off-line at [EMAIL PROTECTED]

Thanks,

Lloyd Good
Circuit Creations
92 Holly Street NW
Calgary AB Canada
T2K 2C8
ph:  (403)282-6445
cell:(403)710-5077




* Tracking #: 1BA43D22D71ECD4592452DA407EE6CC24BCB6244
*


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



[PEDA] DXP Variants / version control

2002-08-26 Thread Peder K. Hellegaard




[PEDA] (No Subject)

2002-08-26 Thread Anand Kulkarni

Hi all,
I am currently designing a printed circuit board in which the main part is a XILINX 
FPGA.

What I want to know is:  (please read on)

Upon  completion of  the schematics of the board design
do PCB designers  do anything to verify the correctness of the board 
schematics like simulation ? 

Especially where something complex like a XILINX FPGA is involved. 

---OR

do PCB designers simply trust the correctness of the schematic and proceed with  PCB 
placement and routing ?

In short ,is any board-level simulation (involving SPICE models or something similar) 
done before fabrication of the board ?

please do reply,
thanks and regards,

Anand Kulkarni



___
Communicate with others using Lycos Mail for FREE!
http://mail.lycos.com 


* Tracking #: F6DD7356D13FD44894E20B0C64260132B0F040F4
*


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



Re: [PEDA] (No Subject)

2002-08-26 Thread Tony Karavidas

Yes, you trust the schematic and its netlist. If it is wrong, then the
person that did the schematic screwed up. Now of course you can tell
them you spotted something, like a power pin that was floating because
they were using hidden pins and forgot to connect the right name, but
don't fix it for them. Alert them and have them drive the fix.



 -Original Message-
 From: Anand Kulkarni [mailto:[EMAIL PROTECTED]] 
 Sent: Monday, August 26, 2002 1:22 PM
 To: PROTEL USER Group
 Subject: [PEDA] (No Subject)
 
 
 Hi all,
 I am currently designing a printed circuit board in which the 
 main part is a XILINX FPGA.
 
 What I want to know is:  (please read on)
 
 Upon  completion of  the schematics of the board design
 do PCB designers  do anything to verify the correctness of the board 
 schematics like simulation ? 
 
 Especially where something complex like a XILINX FPGA is involved. 
 
 ---OR
 
 do PCB designers simply trust the correctness of the 
 schematic and proceed with  PCB placement and routing ?
 
 In short ,is any board-level simulation (involving SPICE 
 models or something similar) done before fabrication of the board ?
 
 please do reply,
 thanks and regards,
 
 Anand Kulkarni
 
 
 
 ___
 Communicate with others using Lycos Mail for FREE! 
http://mail.lycos.com 


* Tracking #: F6DD7356D13FD44894E20B0C64260132B0F040F4
*


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



[PEDA] Fw: Protel 99SE license

2002-08-26 Thread JaMi Smith

LLoyd,

I tried posting direct to you at your [EMAIL PROTECTED] , twice,
and it bombed both tomes with the following fatal error message:

The original message was received at Mon, 26 Aug 2002 16:18:03 -0400
from xxx.xxx.xxx.xxx pacbell.net [xxx.xxx.xxx.xxx]

   - The following addresses had permanent fatal errors -
[EMAIL PROTECTED]

   - Transcript of session follows -
550 5.1.2 [EMAIL PROTECTED]... Host unknown (Name server:
hansel.ugly-duckling.net.: host not found)

I am replying here in the hope that you get this.

JaMi

PS. I xed out my direct dsl address in the above message.

* * * * * * * * * *

- Original Message -
From: JaMi Smith [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Cc: JaMi Smith [EMAIL PROTECTED]
Sent: Monday, August 26, 2002 1:18 PM
Subject: Re: [PEDA] Protel 99SE license


 Lloyd,

 Considering the reception (or lack thereof) of DXP, It is entirely
possible
 that your friend could get P99SE direct from Altium at a substantial
 discount if he was willing to purchase it and forego an ATS License and
free
 Upgrade to DXP.

 Actually, I wouldn't be supprised to find that Altium may continue to sell
 Protel 99 SE, simply because many people do not want to buy into the DXP
 Product since it appears to actually be unproven as a product and have
a
 number of unresolved problems, etc., etc., and also appears to require a
 substantial adaption / learning curve.

 I would check with Altium first, and if you strike out there, I would
 suggest checking with Mr. Lomax, as he may have a license or two for sale.

 JaMi Smith

 -- -- - - -  - - - -- --   -- - -  -   -- - --  -- - --  -- -  - - -  -- -
 -- -  - - --

 - Original Message -
 From: Lloyd Good [EMAIL PROTECTED]
 To: Protel EDA Forum [EMAIL PROTECTED]
 Sent: Monday, August 26, 2002 9:30 AM
 Subject: [PEDA] Protel 99SE license


 
  Hello all,
  I have a client that is looking for a Protel 99SE license. Does anyone
 have
  one for sale. Please email me off-line at [EMAIL PROTECTED]
 
  Thanks,
 
  Lloyd Good
  Circuit Creations
  92 Holly Street NW
  Calgary AB Canada
  T2K 2C8
  ph:  (403)282-6445
  cell:(403)710-5077
 
 
 
  
  * Tracking #: 1BA43D22D71ECD4592452DA407EE6CC24BCB6244
  *
  


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



Re: [PEDA] (No Subject)

2002-08-26 Thread Brad Velander

Anand,
you are asking questions to which the answers shall be as varied as
the number of respondents. The answers will of course be based upon the
environments where the designers work. Some designers are actually EEs and
they do all the design including PCBs, some are EEs but only do the circuit
design and turn the PCB over to the PCB layout person, some are only PCB
designers that don't touch schematics at all. See the problem with an answer
to your question. There is no one answer. The responsibility will be as
equally diverse as the particular environments.

In our shop, I am the PCB designer. I am also responsible for CAD
library generation/maintenance, schematic entry, BOM generation and PCB
design. However I do not design the circuits. The Engineers are responsible
for the overall design. Our circuits are microwave circuits so Protel is
useless for simulation or other integrity verification.

Sincerely,
Brad Velander.

Lead PCB Designer
Norsat International Inc.
Microwave Products
Tel   (604) 292-9089 (direct line)
Fax  (604) 292-9010
email: [EMAIL PROTECTED]
http://www.norsat.com
Norsat's Microwave Products Division has now achieved ISO 9001:2000
certification 



 -Original Message-
 From: Anand Kulkarni [mailto:[EMAIL PROTECTED]]
 Sent: Monday, August 26, 2002 1:22 PM
 To: PROTEL USER Group
 Subject: [PEDA] (No Subject)
 
 
 Hi all,
 I am currently designing a printed circuit board in which the 
 main part is a XILINX FPGA.
 
 What I want to know is:  (please read on)
 
 Upon  completion of  the schematics of the board design
 do PCB designers  do anything to verify the correctness of the board 
 schematics like simulation ? 
 
 Especially where something complex like a XILINX FPGA is involved. 
 
 ---OR
 
 do PCB designers simply trust the correctness of the 
 schematic and proceed with  PCB placement and routing ?
 
 In short ,is any board-level simulation (involving SPICE 
 models or something similar) done before fabrication of the board ?
 
 please do reply,
 thanks and regards,
 
 Anand Kulkarni


* Tracking #: AF1C1D938208214FBB19C969E099D53FCF4343EC
*


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



Re: [PEDA] (No Subject)

2002-08-26 Thread Robison Michael R CNIN

Anand asked:
I am currently designing a printed circuit board in which the main part is a
XILINX FPGA.
What I want to know is:  (please read on)
Upon  completion of  the schematics of the board design
do PCB designers  do anything to verify the correctness of the board 
schematics like simulation ? 
Especially where something complex like a XILINX FPGA is involved. 
do PCB designers simply trust the correctness of the schematic and proceed
with  PCB placement and routing ?
In short ,is any board-level simulation (involving SPICE models or something
similar) done before fabrication of the board ?

***
Hi Anand,

Although we do perform simulation on occasions, its not really 
to verify copper path connections noted in a schematic, but to 
ascertain that component parameters perform as advertised, or,
on the PCB. to examine parasitic or distributed-circuit charac-
teristics.

And I'll be honest with you... I do NOT do simulations for the
physical layout of our boards.   I follow good design rules
when laying the board out, such as keeping traces far enough
apart to prevent too great of crosstalk, using controlled im-
pedances when necessary, and keeping ground and power planes
close together.  I am not saying that you should not perform
this sort of simulation.  We do very small limited runs.  This
allows me to add another couple layers if I want to spread
things out.  In a big production run this would arbitrarily
introduce unacceptable added costs, and the price for an
effective simulation would probably be worthwhile.

good luck, miker


* Tracking #: 06D2151CAADF5C48870F208E48FFABC7CDE708BE
*


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



Re: [PEDA] Tearing my hair out with DXP

2002-08-26 Thread Nick Martin

Hi Tim,

The drawing routines for this were overlooked in the first release but will
be updated to correctly draw the overlapping style in SP1. My apologies
for the confusion.

Best Regards,

Nick Martin
Joint CEO, Altium

- Original Message -
From: Tim Hutcheson [EMAIL PROTECTED]
To: peda [EMAIL PROTECTED]
Sent: Tuesday, August 27, 2002 1:03 AM
Subject: [PEDA] Tearing my hair out with DXP


 All,

 For those that might be trying the multichannel feature of DXP:  I just
 learned the hard way (hours of fiddling) about a few confusing things with
 the method.  The little bit of documentation in the manual, Introducing
 Protel DXP, creates a couple of misunderstandings that I could only
resolve
 by taking the 8-channel Mixer example apart.  So here are some lasting
 impressions from the experience, in the order of points presented in the
 manual.

 First, when you place a Sheet Symbol and add the Repeat(CIN,1,8) style
 Designator, DXP DOES NOT create a Sheet Symbol that looks like the one in
 the Mixer example with multiple overlapping sheets.  These appear to be
just
 artwork added to the mixer schematic as rectangles for dramatic effect
and
 the example would better without them.  I know it is a minor point, but
when
 other things don't work as they probably will not, it left me thinking
that
 I hadn't declared something correctly to invoke the full power of the
 repeat.  True, it is shown correctly in the manual diagram on page 6-2 but
I
 was hard to convince (for about an hour or so, much later, when trying to
 get all to work by generating a netlist without error).

 Second, the technique of representing the bus coming out of a Sheet
Entry
 symbol that has a Repeat as a single wire which then connects directly to
a
 bus, (so that individual signals can be picked out), is so crucial but not
 obvious looking at the example diagram on page 6-2.  There is only a small
 stub of bus wire connected to the Headphone wire in the picture and I
 overlooked that for quite a while.

 Third, it is possible to have the designators in the sheets incremented
 sequentially as _1, _2, ..._n, by simply leaving out the first parameter
of
 the Designator Repeat() statement.  I think this will mean that the
 technique of repeating channels on the circuit layout (assuming you route
by
 hand) by simply copying the first one multiple times should still produce
 the same orderly incrementation of the designator in each channel.  But I
 haven't verified this yet.

 .end - of - rant.

 regards,

 Tim Hutcheson
 Research Associate
 Institute for Human and Machine Cognition
 University of West Florida
 40 S. Alcaniz St.
 Pensacola, FL 32501
 USA
 805-202-4461


 
 * Tracking #: 471375E444F874419B41CCFBFC51A3346C4D8132
 *
 


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



Re: [PEDA] (No Subject)

2002-08-26 Thread Jon Elson

Anand Kulkarni wrote:

 Hi all,
 I am currently designing a printed circuit board in which the main part is a XILINX 
FPGA.

 What I want to know is:  (please read on)

 Upon  completion of  the schematics of the board design
 do PCB designers  do anything to verify the correctness of the board
 schematics like simulation ?

 Especially where something complex like a XILINX FPGA is involved.

 ---OR

 do PCB designers simply trust the correctness of the schematic and proceed with  PCB 
placement and routing ?

 In short ,is any board-level simulation (involving SPICE models or something 
similar) done before fabrication of the board ?

Generally, I think the answer is no.  There may be some need to simulate critical
traces, such as high-speed clocks, to evaluate need for termination, and similar
things.  A complete analog  simulation of a complex digital circuit including
the FPGA is beyond the capacity of todays computers, unless you have a HUGE
budget, and a few months of real supercomputer time available.

PCB designers generally are not involved in the verification of the circuitry,
especially what is programmed into an FPGA.

My personal approach to this is to possibly simulate a generalized case of a
single signal, like one bit of a data bus, or possibly 2 bits to check cross-talk.
Once you know the signal quality of a representative signal, and have varied
some of the parameters to see over what range of trace lengths, for instance, proper
operation is obtained, you can then apply this as a 'rule' against all the signals
of the same character.

Some of the people designing Pentium motherboards are supposed to be
using lots of simulation, but I have no idea what tools they are using.
I think it is pretty clear they are NOT using Protel99 for this.

Jon



* Tracking #: 5BBAE86F2A670642A7A4D94DB8F0729E3A643CF8
*


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