Hi Ekkehard,
<blockquote id="quote"><font size="1" face="Verdana, Arial, Helvetica" id="quote">Zitat:<hr height="1" noshade id="quote">die Firma Orion England hat ja seit einiger Zeit ein Zygo Interferometer.
Wie ich hörte, sollen seit dem die Optiken von Orion erheblich besser sein.
Wenn´s so ist, hat sich die Anschaffung ja auch qualitativ bei Orion niedergeschlagen.
Meine Frage:
Ich habe nun mehrfach gehört daß ein Zygo Interferometer-Protokoll für den Käufer eine feine sache sein soll, da man sich darauf angeblich 100%ig verlassen kann - Manipulation ausgeschlossen.
Stimmt das?<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">
hier die Antwort von Barry Pemberton persönlich (mit der ausdrücklichen Erlaubnis der Weiterleitung [s. ganz unten]):
I am not at all good at the German language and automatic translators are
unreliable where technical points are concerned. This is the main reason why
I take little notice of German notice boards. I am sure they contain very
interesting subjects but, I am also very aware that even with a modest
translation service, there are some very arrogant members who try to stir up
trouble with very little knowledge behind their voices.
Allow me to tell you exactly why we bought (over 100,000 Euros) a GPI Zygo:
We previously had a Mk4 Zygo which is a good instrument but, it has two
massive problems:
It is virtually impossible to test parabolas accurately. The reason for
this is because parabolas with the Zygo Mk4, Mk3 and earlier models must be
used with a flat on a double pass to produce an autocollimator. This has
with it one large problem, possibly two.
1. It is very easy indeed to introduce astigmatism into the test due to
mis-collimation of the two optical components. All optics have degrees of
astigmatism but, it must be measurable in the optic under test and not the
mis-alignment of the system being tested. I will give you an example. A Zygo
Mk3 is used in autocollimation with a flat with a central hole, we will
assume here, the flat is absolutely perfect. The mirror under test will have
some astigmatism, this is guaranteed, no matter who made it. When the
optician is aligning the system he will do so to try and have as little
astigmatism visible in the test because he assumes the astigmatism is due to
miscollimation. It is possible he may be able to virtually remove every
trace of astigmatism from the test by aligning the flat and mirror in a
specific way. If he now measures this mirror under test, he assumes
incorrectly that the mirror has no astigmatism which is wrong. What he has
done is to 'remove' it by adding astigmatism by a similar amount in an
opposite axis by mis-alignment of the optics.
What he can also do is remove all astigmatism from the calculations,
assuming the residual astigmatism is purely because he has not correctly
aligned the components. This is again wrong.
In both cases it is impossible to measure the astigmatism in the mirror
because of opposite compensation astigmatism being intentionally added by
mis-alignment and, in the second instance, by intentionally removing all
astigmatism in the test which not only removes astigmatism through
mis-alignment but, in the mirror also.
2. To measure a mirror's wavefront accurately many averages have to be
used. With a Mk3 or Mk4, it is possible, under extremely favourable
conditions, to average sometimes up to 10. We only ever managed 5 averages
due to atmospheric disturbances and vibration. It takes, if I remember
correctly, about 10 minutes to carry out 5 averages. Unless many averages
are used, these two effects, air disturbance (temperature differences) and
vibration, can ruin a test.
So, what will a GPI do that a MK3 or MK4 cannot possibly do?
Well, to refer to #1 above. Because of the software and engineering involved
in a GPI, it can test parabolas, spheres, hyperbolas and ellipsoids at their
radius of curvature with no other optical component involved. This means any
astigmatism present in the test is 100% in the mirror and can be measured
consistently and accurately, again and again. We feed the theoretical value
of a 100% correct mirror into the Zygo and subtract all the Zernike values
which were measured by the Zygo. The difference is the exact sum of the
errors in the mirror's wavefront.
Referring to #2. The GPI can average thousands of readings from one mirror.
As an example, it measures around 150 sets of values every minute, so in the
same 10 minute period where a MK3 or Mk 4 would measure between about 5-10,
the GPI will measure about 1,500 and average them.
These 2 points above are the main reasons why we bought the GPI, there are
others but they are less significant.
What has it done for our quality?
An amazing amount. Because of its speed we put a great deal of reliability
in its results. It allows us to look at specific areas on the mirror's
wavefront and then see how high or low the corresponding mirror's surface is
to cause this error(s) on the wavefront. When a mirror of say Lambda/5 is on
test, we can see all the offending high areas on the wavefront which cause
this figure. We then draw with a spirit pen on the mirror's surface when
still on test, around these areas. The optician then takes the mirror back
to the optical laboratory and very carefully polishes by hand these 'high'
areas, within the spirit pen areas. He then puts the mirror back on test in
the same orientation 'North to North' on the test stand and re-tests with
the Zygo to see what the improvement is.
With this method we are able to figure large mirrors to incredible
accuracies, the time involved being obviously high and the price similarly
so. We do this regularly for mirrors with higher accuracies tan Lambda/5 and
quite regularly produce 300mm f4 mirrors, as an example, to lambda/12 on the
wavefront.
As you can see, the GPI is a massive leap over all previous Zygo
interferometers, especially in software, speed and repeatability.
To comment on the stupid people who suggest we 'adjust' our reports. What is
the point of spending 100,000 Euros to fabricate false results when we could
do the same with a 15,000 Euros old Mk3 or Mk4 Zygo. If those people are so
stupid as not able to see how ridiculous their remarks are, they really are
not fit to be allowed to partake in a sensible thread on optics.
I hope you found these brief notes useful and have given you a little more
insight into our quality control and testing techniques. Please contact me
if you require any further information.
Regards
Barry
You may share it with anyone you wish. No one should be afraid to hear the
truth. Please let me know, in English, some of the comments you get.
Regards
Barry
Orion Optics
Units 21 & 22 Third Avenue
Crewe
Cheshire
CW1 6XU
England
00 44 1270 500089 Phone
00 44 1270 251213 Fax
http://www.orionoptics.co.uk
cs, m.
