Electronics-Related.com
Forums

Stability of older Orcad/PSpice combos?

Started by Joerg April 24, 2011
Hello Folks,

On my PC this software crashes often and hard and needs too many
re-installs. It slows down one project to a crawl here and that needs to
get done. I could use older versions as long as my client can read the
files into 16.3. Question:

In your experience, what was a relatively stable (as in "not many
crashes") version I could try to go back to? The license should allow a
downgrade. The support server won't let me in right now but I'll bug
them about that later.

It'll be a bit of a pain since I won't be able to revisit circuits I
already completed with 16.3 but I may not have a choice. Or maybe
they'll let me use both old and new, the suppoort folks are very friendly.

-- 
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
On Sun, 24 Apr 2011 08:18:01 -0700, Joerg <invalid@invalid.invalid> wrote:

>Hello Folks, > >On my PC this software crashes often and hard and needs too many >re-installs. It slows down one project to a crawl here and that needs to >get done. I could use older versions as long as my client can read the >files into 16.3. Question:
He can read older files into 16.3 but you won't be able to read his at all if you're too many levels back.
>In your experience, what was a relatively stable (as in "not many >crashes") version I could try to go back to? The license should allow a >downgrade. The support server won't let me in right now but I'll bug >them about that later.
Good luck! Stable and OrCAD are mutually exclusive. IOW, you *can't* go back that far.
>It'll be a bit of a pain since I won't be able to revisit circuits I >already completed with 16.3 but I may not have a choice. Or maybe >they'll let me use both old and new, the suppoort folks are very friendly.
16.3 can output older files, but not old enough for you to find a "stable" OrCAD (the thought makes me laugh). BTW, I've found that 16.3 *is* more stable than others, at least recent ones. It's still a POS, but less of one. That said, Cadence isn't happy about supporting older versions. You're on your own. We have a single-user 15.7 OrCAD + Layout + P-Spice seat. When they installed the floating 16.3 license that copy stopped working (so no P-Spice and that copy of Layout stopped working). Cadence didn't care much that we lost P-Spice. It was out of service so not their problem.
On Sun, 24 Apr 2011 08:18:01 -0700, Joerg <invalid@invalid.invalid>
wrote:

>Hello Folks, > >On my PC this software crashes often and hard and needs too many >re-installs. It slows down one project to a crawl here and that needs to >get done. I could use older versions as long as my client can read the >files into 16.3. Question: > >In your experience, what was a relatively stable (as in "not many >crashes") version I could try to go back to? The license should allow a >downgrade. The support server won't let me in right now but I'll bug >them about that later. > >It'll be a bit of a pain since I won't be able to revisit circuits I >already completed with 16.3 but I may not have a choice. Or maybe >they'll let me use both old and new, the suppoort folks are very friendly.
Jeorg, You should be able to use an older version (with caveats!) with the new licenses. The older the version, the more likely there was some change in the licensing that would break it, though. Anything I can do to help? I know a thing or two about the software... ;-) Charlie
On Sun, 24 Apr 2011 09:17:33 -0700, Charlie E. <edmondson@ieee.org>
wrote:

>On Sun, 24 Apr 2011 08:18:01 -0700, Joerg <invalid@invalid.invalid> >wrote: > >>Hello Folks, >> >>On my PC this software crashes often and hard and needs too many >>re-installs. It slows down one project to a crawl here and that needs to >>get done. I could use older versions as long as my client can read the >>files into 16.3. Question: >> >>In your experience, what was a relatively stable (as in "not many >>crashes") version I could try to go back to? The license should allow a >>downgrade. The support server won't let me in right now but I'll bug >>them about that later. >> >>It'll be a bit of a pain since I won't be able to revisit circuits I >>already completed with 16.3 but I may not have a choice. Or maybe >>they'll let me use both old and new, the suppoort folks are very friendly. > >Jeorg, >You should be able to use an older version (with caveats!) with the >new licenses. The older the version, the more likely there was some >change in the licensing that would break it, though. Anything I can >do to help? I know a thing or two about the software... ;-) > >Charlie
I believe Joerg is using Crapture. That is a guarantee of "lossy" crashes :-( ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | Remember: Once you go over the hill, you pick up speed
Joerg wrote:

> Hello Folks, > > On my PC this software crashes often and hard and needs too many > re-installs. It slows down one project to a crawl here and that needs to > get done. I could use older versions as long as my client can read the > files into 16.3. Question: > > In your experience, what was a relatively stable (as in "not many > crashes") version I could try to go back to? The license should allow a > downgrade. The support server won't let me in right now but I'll bug > them about that later. > > It'll be a bit of a pain since I won't be able to revisit circuits I > already completed with 16.3 but I may not have a choice. Or maybe > they'll let me use both old and new, the suppoort folks are very friendly. >
IT would be nice to know what type of crashes you are getting? Are these actual built in program messages being generated that causes corruption of your work? It so, you could be taxing some function a bit and the software should know how to recovery it or step backwards that is. If you are getting things like invalid pointers or out of range indexes and things like that? I would venture to say that some one is not doing QC very well... If you are getting errors that show coming from System API's, kernel faults and such, the ones that really causes the rest of your machine a reboot requirement? I would say that you have have something wrong with your PC? I've found that bad memory and a flaky HD (swap file) can cause random problems on top of a machine running to hot and maybe was over clocked! It's always that possibility that you are just working it hard and the program is reaching it's bows with the allocated memory requirements and things like that. I've seen where sloppy use of memory handling can cause the memory pool to get fragmented and some where along the line, windows will just not be able to allocate a piece of memory for the code with no safety checks in the code. Are you using a true Win32 version or some older 16 bit version? It does make a big difference. Jamie
Jim Thompson wrote:
> On Sun, 24 Apr 2011 09:17:33 -0700, Charlie E. <edmondson@ieee.org> > wrote: > >> On Sun, 24 Apr 2011 08:18:01 -0700, Joerg <invalid@invalid.invalid> >> wrote: >> >>> Hello Folks, >>> >>> On my PC this software crashes often and hard and needs too many >>> re-installs. It slows down one project to a crawl here and that needs to >>> get done. I could use older versions as long as my client can read the >>> files into 16.3. Question: >>> >>> In your experience, what was a relatively stable (as in "not many >>> crashes") version I could try to go back to? The license should allow a >>> downgrade. The support server won't let me in right now but I'll bug >>> them about that later. >>> >>> It'll be a bit of a pain since I won't be able to revisit circuits I >>> already completed with 16.3 but I may not have a choice. Or maybe >>> they'll let me use both old and new, the suppoort folks are very friendly. >> Jeorg, >> You should be able to use an older version (with caveats!) with the >> new licenses. The older the version, the more likely there was some >> change in the licensing that would break it, though. Anything I can >> do to help? I know a thing or two about the software... ;-) >>
Thanks, Charlie. I might take you up on that.
>> Charlie > > I believe Joerg is using Crapture. That is a guarantee of "lossy" > crashes :-( >
Well, yeah, that's part of 16.3 so I am kind of forced to use Capture. That's why I wanted to know, how far back can I safely go without my client losing the ability to read in my stuff? -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
Jamie wrote:
> Joerg wrote: > >> Hello Folks, >> >> On my PC this software crashes often and hard and needs too many >> re-installs. It slows down one project to a crawl here and that needs to >> get done. I could use older versions as long as my client can read the >> files into 16.3. Question: >> >> In your experience, what was a relatively stable (as in "not many >> crashes") version I could try to go back to? The license should allow a >> downgrade. The support server won't let me in right now but I'll bug >> them about that later. >> >> It'll be a bit of a pain since I won't be able to revisit circuits I >> already completed with 16.3 but I may not have a choice. Or maybe >> they'll let me use both old and new, the suppoort folks are very >> friendly. >> > IT would be nice to know what type of crashes you are getting? > > Are these actual built in program messages being generated that causes > corruption of your work? It so, you could be taxing some function a bit > and the software should know how to recovery it or step backwards that is. > If you are getting things like invalid pointers or out of range > indexes and things like that? I would venture to say that some one is > not doing QC very well... > > If you are getting errors that show coming from System API's, kernel > faults and such, the ones that really causes the rest of your machine a > reboot requirement? I would say that you have have something wrong with > your PC? I've found that bad memory and a flaky HD (swap file) can cause > random problems on top of a machine running to hot and maybe was > over clocked! > > It's always that possibility that you are just working it hard and the > program is reaching it's bows with the allocated memory requirements and > things like that. I've seen where sloppy use of memory handling can > cause the memory pool to get fragmented and some where along the line, > windows will just not be able to allocate a piece of memory for the code > with no safety checks in the code. >
Yes, I was working it quite hard. I don't know much about software but most or probably all of the smoke and spatters seem to come from within the app (Orcad) itself. Messages such as "The PSPice COM wrapper error has occurred" cannot possibly come from the Windows OS. Then "This application has quit unexpectedly. To help Cadence ..." is also pretty cliear about the origin. Then there's what I have now, it just refuses to do certain simple things, like opening a project. Instead of opening or at least displaying some error message Orcad just locks up hard.
> Are you using a true Win32 version or some older 16 bit version? It > does make a big difference. >
There is only one version and it's supposed to run on Windows XP. Amnd it does run. For a few days at a time :-( -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
krw@att.bizzzzzzzzzzzz wrote:
> On Sun, 24 Apr 2011 08:18:01 -0700, Joerg <invalid@invalid.invalid> wrote: > >> Hello Folks, >> >> On my PC this software crashes often and hard and needs too many >> re-installs. It slows down one project to a crawl here and that needs to >> get done. I could use older versions as long as my client can read the >> files into 16.3. Question: > > He can read older files into 16.3 but you won't be able to read his at all if > you're too many levels back. > >> In your experience, what was a relatively stable (as in "not many >> crashes") version I could try to go back to? The license should allow a >> downgrade. The support server won't let me in right now but I'll bug >> them about that later. > > Good luck! Stable and OrCAD are mutually exclusive. IOW, you *can't* go back > that far. >
Well, if I just knew how far "that far" is :-) But of course I don't have too high hopes because I did manage to thoroughly crash older version (I believe 9.something) at a client. They weren't too fussed because I wasn't the only one. [...] -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
Joerg wrote:

> Jamie wrote: > >>Joerg wrote: >> >> >>>Hello Folks, >>> >>>On my PC this software crashes often and hard and needs too many >>>re-installs. It slows down one project to a crawl here and that needs to >>>get done. I could use older versions as long as my client can read the >>>files into 16.3. Question: >>> >>>In your experience, what was a relatively stable (as in "not many >>>crashes") version I could try to go back to? The license should allow a >>>downgrade. The support server won't let me in right now but I'll bug >>>them about that later. >>> >>>It'll be a bit of a pain since I won't be able to revisit circuits I >>>already completed with 16.3 but I may not have a choice. Or maybe >>>they'll let me use both old and new, the suppoort folks are very >>>friendly. >>> >> >> IT would be nice to know what type of crashes you are getting? >> >> Are these actual built in program messages being generated that causes >>corruption of your work? It so, you could be taxing some function a bit >> and the software should know how to recovery it or step backwards that is. >> If you are getting things like invalid pointers or out of range >>indexes and things like that? I would venture to say that some one is >>not doing QC very well... >> >> If you are getting errors that show coming from System API's, kernel >>faults and such, the ones that really causes the rest of your machine a >>reboot requirement? I would say that you have have something wrong with >>your PC? I've found that bad memory and a flaky HD (swap file) can cause >>random problems on top of a machine running to hot and maybe was >>over clocked! >> >> It's always that possibility that you are just working it hard and the >>program is reaching it's bows with the allocated memory requirements and >>things like that. I've seen where sloppy use of memory handling can >>cause the memory pool to get fragmented and some where along the line, >>windows will just not be able to allocate a piece of memory for the code >>with no safety checks in the code. >> > > > Yes, I was working it quite hard. I don't know much about software but > most or probably all of the smoke and spatters seem to come from within > the app (Orcad) itself. Messages such as "The PSPice COM wrapper error > has occurred" cannot possibly come from the Windows OS. Then "This > application has quit unexpectedly. To help Cadence ..." is also pretty > cliear about the origin. Then there's what I have now, it just refuses > to do certain simple things, like opening a project. Instead of opening > or at least displaying some error message Orcad just locks up hard. > > > >> Are you using a true Win32 version or some older 16 bit version? It >>does make a big difference. >> > > > There is only one version and it's supposed to run on Windows XP. Amnd > it does run. For a few days at a time :-( >
I call that junk. If you are calling com objects, there is always that chance that an error will happen or some unexpected early return from the COM object. CHecking the status of the object on each return should be done, if it's available. And calling code to outside boundaries should be wrapped in a TRY+EXCEPT block so that the application can gain control and do proper clean up before destroying your work. More than likely the lock up from project opening is the end results of project files getting corrupted due to the uncontrolled termination of the program. Hope you had back ups.. Trying to start a new project would indicate this to be a fact, if the new project succeeds in creating itself. If not, then you must have some left over config file that got corrupted. Oh well, it sucks. Personally, I don't like using third party COM files unless they are well supported to help out the users of it. Did OrCad take some change of ownership/dev team at some recent point? If so, it could explain a lot about it being buggy. I've always found when a new owner puts a paint job on an old shell, trouble is brewing! Jamie
Jamie wrote:
> Joerg wrote: > >> Jamie wrote: >>
[...]
>>> Are you using a true Win32 version or some older 16 bit version? It >>> does make a big difference. >>> >> >> >> There is only one version and it's supposed to run on Windows XP. Amnd >> it does run. For a few days at a time :-( >> > I call that junk. If you are calling com objects, there is always that > chance that an error will happen or some unexpected early return from > the COM object. CHecking the status of the object on each return should > be done, if it's available. And calling code to outside boundaries > should be wrapped in a TRY+EXCEPT block so that the application can gain > control and do proper clean up before destroying your work. > > More than likely the lock up from project opening is the end results > of project files getting corrupted due to the uncontrolled termination > of the program. Hope you had back ups.. >
Yes, I do. Of course not from the last few hours so some is lost but not too bad. What is a problem is that this dreaded CAD is out of commission so often. Currently doing an uninstall. Again. Re-install takes a long time because some (to me nonsensical) help index build takes place. Then when you click help it does some more indexing. Beats me why.
> Trying to start a new project would indicate this to be a fact, if the > new project succeeds in creating itself. If not, then you must have some > left over config file that got corrupted. >
I didn't get that far because Orcad completely self-destructed and needed a re-install.
> Oh well, it sucks. > > Personally, I don't like using third party COM files unless they are > well supported to help out the users of it. > > Did OrCad take some change of ownership/dev team at some recent > point? If so, it could explain a lot about it being buggy. I've always > found when a new owner puts a paint job on an old shell, trouble is > brewing! >
That was a long time ago. Orcad was bought by Cadence. By now my impression is that this might not have been such a good thing to happen. Oh well. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.