ZEMAX tolerancing non-sequential mode
chibitul
distribution. Fine, but how do I do a tolerance study, it seems to be
available only in sequential mode. I just started using ZEMAX a few
days ago, perhaps I am missing something obvious? Thanks guys. oh, no
flames, please. ZEMAX is what I have to use. I also used OSLO in a
previous job.
Rob
"chibitul" <chib...@sbcglobal.net> wrote in message
news:1149629466.7...@h76g2000cwa.googlegroups.com...
Non sequential tolerancing is possible, using TNPS and TNPA for instance.
When you just started using Zemax, I would not recommend tolerancing by the
way...
And Zemax has support too when you pay for it.
Rob
chibitul
Rob wrote:
> Non sequential tolerancing is possible, using TNPS and TNPA for instance.
> When you just started using Zemax, I would not recommend tolerancing by the
> way...
why? could you please explain?
> And Zemax has support too when you pay for it.
yes, we padi for it, I have an USB key, and I was gonna check the
details of the license and support, like I said, I didn't have a chance
to look into all details very carefully. thanks!
Jim Klein
The fact is, ZEMAX is the correct choice for many applications. No
flames. Ken has worked very hard and has spent many more hours late
than I have to write his program.
For the less well $ endowed, I have an alternative that works very
well for most design applications, many specialized analysis
applications and also has good support and full source code access.
Now we can stop waiting for a nasty post from me which will not
arrive. :-)
James E. Klein
james...@earthlink.net
Engineering Calculations
http://www.ecalculations.com
ecalcu...@ecalculations.com
Engineering Calculations is the home of
the KDP-2 Optical Design Program
for Windows and (soon) MAC OSX
Free KDP-2 (DEMO) downloadable!
1-818-507-5706 (Voice and Fax)
Skywise
news:vmme82tscdmlh7l94...@4ax.com:
<Snipola>
> Now we can stop waiting for a nasty post from me which will not
> arrive. :-)
So Jim, how much was the bribe? or the threatened suit? ;)
Brian
--
http://www.skywise711.com - Lasers, Seismology, Astronomy, Skepticism
Seismic FAQ: http://www.skywise711.com/SeismicFAQ/SeismicFAQ.html
Quake "predictions": http://www.skywise711.com/quakes/EQDB/index.html
Sed quis custodiet ipsos Custodes?
Rob
Why I would not recommend this to someone using Zemax just for a few days:
The tolerancing is quite straightforward using Zemax, but without someone
explaining the details (or divning into it yourself for a significant period
of time) I think it is hard to get grip of. Of course, I do not know your
background or if you have performed tolerancing using another ray-tracing
software package (code-V, oslo, whatever).
The tolerancing will spit out number even if you don't understand what is
exactly is doing, which is a dangerous situation...
Regards,
Rob
Helpful person
Rob wrote:
>
> Why I would not recommend this to someone using Zemax just for a few days:
> The tolerancing is quite straightforward using Zemax, but without someone
> explaining the details (or divning into it yourself for a significant period
> of time) I think it is hard to get grip of. Of course, I do not know your
> background or if you have performed tolerancing using another ray-tracing
> software package (code-V, oslo, whatever).
> The tolerancing will spit out number even if you don't understand what is
> exactly is doing, which is a dangerous situation...
>
> Regards,
> Rob
I am far from a fan of Zemax or of Focus Software. However,
tolerancing is one of the few areas of the program that does work
extremely well.
Forget the numerical analysis. Use the Monte Carlo option with update
graphics. You can then see exactly how your system will change for the
parameters that you input.
Please visit my web site at www.richardfisher.com
Jim Klein
>Jim Klein <james...@earthlink.net> wrote in
>news:vmme82tscdmlh7l94...@4ax.com:
>
><Snipola>
>> Now we can stop waiting for a nasty post from me which will not
>> arrive. :-)
>
>So Jim, how much was the bribe? or the threatened suit? ;)
>
>Brian
I grew up. Ken won. I still provide to the poor.
Rob
"Helpful person" <rrl...@yahoo.com> wrote in message
news:1149768434.7...@u72g2000cwu.googlegroups.com...
I am a fan of Zemax, if not only because that is the package I use most of
the time :-)
The monte carlo option just calculates statistics using the parameters
filled in in the toleracne data editor. Useful sure, but not by itself...
sensitivity and inverse tolerancing and MC are all needed for the full
picture.
Rob
Helpful person
Rob wrote:
>
> I am a fan of Zemax, if not only because that is the package I use most of
> the time :-)
> The monte carlo option just calculates statistics using the parameters
> filled in in the toleracne data editor. Useful sure, but not by itself...
> sensitivity and inverse tolerancing and MC are all needed for the full
> picture.
>
> Rob
Not true. I do not have Zemax loaded at pesent so this is from memory.
If you open the TOL box there is a check box that says something like
"Update MC graphics". If checked, this will upgrade all open graphics
windows, the MC decks' graphs all being on the same window. Difficult
to explain. Try it.
(If you are a fan of Zemax you are either not designing complex
systems, are not an experienced lens designe, are a reformed Jim Klein
or have been very lucky.)
Rob
I do have Zemax opened and I am performing tolerance analyses for some years
with it now, so I think I am quite experienced with it. The best overall
advise: Know exactly what you and Zemax are doing before jumping to
conclusions. In my opinion (and there are more ways to Rome...) using "only"
the monte carlo analysis is insufficient, among others because it changes
all parameters witthin the statistics defined. You get no clue on what
parameters are more sensitive than others. Of course, the opposite is also
true: You cannot do a complete analysis without the monte carlo analysis.
I also dare to say that we are designing very complex systems, but for sure
I am not an experienced designer (as far as I can tell only people with >20
years of experience can say that).
I am not praising Zemax in particular to heaven, but I would be quite
helpless without it.
Rob.
Helpful person
Rob wrote:
> "Helpful person" <rrl...@yahoo.com> wrote in message
> news:1149856073....@i39g2000cwa.googlegroups.com...
>
> with it now, so I think I am quite experienced with it. The best overall
> advise: Know exactly what you and Zemax are doing before jumping to
> conclusions. In my opinion (and there are more ways to Rome...) using "only"
> the monte carlo analysis is insufficient, among others because it changes
> all parameters witthin the statistics defined. You get no clue on what
> parameters are more sensitive than others. Of course, the opposite is also
> true: You cannot do a complete analysis without the monte carlo analysis.
>
> I also dare to say that we are designing very complex systems, but for sure
> I am not an experienced designer (as far as I can tell only people with >20
> years of experience can say that).
> I am not praising Zemax in particular to heaven, but I would be quite
> helpless without it.
>
> Rob.
I thoroughly agree with you that one needs to "Know exactly what you
and Zemax are doing before jumping to conclusions.". It's analysis has
too many problems. I also agree that only using Monte Carlo is
dangerous. However, except for the realization that tolerances never
obey nice Gaussian distributions, I believe that it is the best guide
available.
Monte Carlo does not give individual sensitivities of lenses. A guide
to this is seen in the sensitivities analysis. Changing all parameters
at once during a tolerance analysis is essential, as an optical
system's parameters are not orthogonal.
I consider the job of tolerancing an optical system much more complex
(and certainly more time consuming) than designing the system in the
first place. It's the tolerancing that sorts out the men from the
boys.
I admit that I also use Zemax as I own a copy. At present I cannot
justify the expense of buying a new program.
Please visit my web site at www.richardfisher.com
(Contrary to what some people believe, I post my web site not to
generate business but to add a link for search engine bots to
discover.)
Rob
Maybe it helps: The statistical distribution does not need to be normal
distributed in a monte carlo run. The setting can be put to normal, uniform
or parabolic. Looking for the worst-case could be faster in the parabolic,
where the emphasis is on the large end of the tolerance width.
Rob
Helpful person
Rob wrote:
>
> Maybe it helps: The statistical distribution does not need to be normal
> distributed in a monte carlo run. The setting can be put to normal, uniform
> or parabolic. Looking for the worst-case could be faster in the parabolic,
> where the emphasis is on the large end of the tolerance width.
>
> Rob
Whichever distribution you choose will still not be the same as the
real world. Different parts manufacturers have different distributions
depending on their processes. Worst case is not always the most
economical.
Jim Klein
I didn't say I was a fan of ZEMAX.
I'm a fan of Code-V and HEXAGON and I sort of like KDP-2. :-)
Helpful person
> "Helpful person" <rrl...@yahoo.com> wrote:
>
> >
> >(If you are a fan of Zemax you are either not designing complex
> >systems, are not an experienced lens designe, are a reformed Jim Klein
> >or have been very lucky.)
> >
>
>
> I'm a fan of Code-V and HEXAGON and I sort of like KDP-2. :-)
> James E. Klein
> james...@earthlink.net
>
>
Sorry, I didn't mean to put words in your mouth.
Programs I have used and liked are Code V (naturally), SLAMS
(sentimentally, it was the program I used, from Imperial college in the
1970s), HOAD (I had access to the source code and made many additions)
and the DOS version of Synopsys.