TEC 의 사장인 Yuri 가 Yahoo 의 TEC 그룹에 올린 글입니다.
air spaced 와 oil spaced 의 장단점에 대해 논하고 있는데, 아무래도
자기 회사 망원경이 oil spaced 이다 보니 oil 쪽에 더 좋은 평을 하고 있군요.
Once again about an old question, or oil vs. air.
Each design has its own positive and negative sides, but being not so much
different in most discussed area - "color correction", the last one could
be an easy suject for speculation by companies that usually do not involve
in any aspects of optical production, and being this thiw way can easy
announce for their airspaced design super correction of "1/100 of the
conventional apochromats telescope, it has reached the aberration quantity
which is much smaller than the catadioptric system..."; or the other good
sample: "The color correction on the XXX 203mm f/7 will be the best possible
of any existing apochromat design on the market" - we can only smile and
remind that optics on the computer screen has no limitations:
- no streas or bubbles, blanks are uniformly annealed from center to edge,
catalog refractive indexes, no any surface impefections, all surfaces perfectly
polished, etc. - what a hard work of the computer! Is that same easy easy in
real life? Lets see.
Both oiled and airspaced designs could be done with very good or near
prefect correction and not only on the computer screen. What is the difference
in numbers for color correction? - if talking about eye most sensitive wave
length of ~ 532nm (green light), the difference is ZERO, since both designs
optimized for this wave length.
Going to polychromatic Strehl, that counts whole visual range from 430nm to
over 650nm (or slightly wider) the situation becomes a bit better for airspaced
design, but how much is that "a bit": - sorry to say, but it is only 3 - 5% (!).
Could this difference be realized, yes; anything behind - yes again - the glass
measurements and glass internal quality has to be taking in account.
Refractive indexes of the glasses given by manufacturer for standard
measurements are with in +- 3.0 x10E-5, same time each melt of glass may be
different from catalog numbers up to 10 x 10E-4, and variations in one melt
could be a serious issue when going to larger lenses (>150mm dia.)! -
Counting only these will bring to nought "theoretical" 3-5% advantage of
airspaced design.
After being "surprised" for a few times with melt data given by manufacturers,
we do our own tests of refractive indexes of each glass on more precision
level: <1x10E-5, as a result we have the MEASURED correction of
the final objective be very close or identical to design. That not easy but
we have to do this for each new melt and if different - we have to do new
sets of test plates, tooling re-setup, etc. - All of these usually ignored
by most manufacturers, since requires time, special equipment and test
methods. For sure that is ignored by all mass production manufacturers (Asia).
Other aspect in question - coatings. The math is not needed to show
difference in cost of coating of 2 outer surfaces of oiled objective vs. all
6 surfaces of airspased objective.
No math also needed to show the winner in light transmission; light loss on
reflections from 6 surfaces of airspaced design is ~3% in the middle of visual
range and 6% and more on the edges of it - and this is in case if we have
perfect coatings; if it is an average only - the light loss on the on 6 surfaces
on
the blue end of spectrum may exceed >10-20% (deep blue looking coatings),
that will shift color balance and make background yellowish. In the oiled
design the light loss on reflections is only 0.5-0.7 %.
What is next? - Scattering light, on six surfaces three times more than on two -
where is the higher contrast to be? How about accumulated errors on 6
surfaces vs. 2 - for sure 6 surfaces has to be done on more precision level to
reach quality of two outer surfaces of oiled design - that takes more time and
increases the cost of production. Will accumulated errors affect Strehl - why
not - in best case same 3-5% or even more of polychromatic Strehl will be lost.
BTW, the third order spherical aberration of 1/4 PV resulting dramatic Strehl
lowering to 80%! That is the area of attention - the optical surfaces must be
done on highest level possible!
You may see know why we choose and stay with an oiled design in all of our
APOs.
About glass choices, or in the other words, any advantages for CaF2.
Even first three APO160 F8 were done with CaF2, we had a plan to produce
APO160 with ED glass (price difference for blanks for APO140ED vs.
APO160ED is only $400) and be priced for appr. $8000, but we went to
Fluorite. Why? Our work experience and testing the ED glasses and CaF2
blanks in sizes up to 200mm and larger showed us that homogeneity of high
quality single crystals of CaF2 was better than in ED glasses, same time the
refractive index of CaF2 is constant up to 5x10E-6, that means no
measurements are needed; no refractive index variations, as in ED glass
blanks of sizes over 6" diameter. As being said before, CaF2 by being a
crystal has near zero internal light scatter and after all the Fluorite design
gave us possibility to force focal ratio to F7 without compromising color
correction in wide range including blue end - same level in practical
realization on the as in the best airspaced lenses.
Best regards, TEC
air spaced 와 oil spaced 의 장단점에 대해 논하고 있는데, 아무래도
자기 회사 망원경이 oil spaced 이다 보니 oil 쪽에 더 좋은 평을 하고 있군요.
Once again about an old question, or oil vs. air.
Each design has its own positive and negative sides, but being not so much
different in most discussed area - "color correction", the last one could
be an easy suject for speculation by companies that usually do not involve
in any aspects of optical production, and being this thiw way can easy
announce for their airspaced design super correction of "1/100 of the
conventional apochromats telescope, it has reached the aberration quantity
which is much smaller than the catadioptric system..."; or the other good
sample: "The color correction on the XXX 203mm f/7 will be the best possible
of any existing apochromat design on the market" - we can only smile and
remind that optics on the computer screen has no limitations:
- no streas or bubbles, blanks are uniformly annealed from center to edge,
catalog refractive indexes, no any surface impefections, all surfaces perfectly
polished, etc. - what a hard work of the computer! Is that same easy easy in
real life? Lets see.
Both oiled and airspaced designs could be done with very good or near
prefect correction and not only on the computer screen. What is the difference
in numbers for color correction? - if talking about eye most sensitive wave
length of ~ 532nm (green light), the difference is ZERO, since both designs
optimized for this wave length.
Going to polychromatic Strehl, that counts whole visual range from 430nm to
over 650nm (or slightly wider) the situation becomes a bit better for airspaced
design, but how much is that "a bit": - sorry to say, but it is only 3 - 5% (!).
Could this difference be realized, yes; anything behind - yes again - the glass
measurements and glass internal quality has to be taking in account.
Refractive indexes of the glasses given by manufacturer for standard
measurements are with in +- 3.0 x10E-5, same time each melt of glass may be
different from catalog numbers up to 10 x 10E-4, and variations in one melt
could be a serious issue when going to larger lenses (>150mm dia.)! -
Counting only these will bring to nought "theoretical" 3-5% advantage of
airspaced design.
After being "surprised" for a few times with melt data given by manufacturers,
we do our own tests of refractive indexes of each glass on more precision
level: <1x10E-5, as a result we have the MEASURED correction of
the final objective be very close or identical to design. That not easy but
we have to do this for each new melt and if different - we have to do new
sets of test plates, tooling re-setup, etc. - All of these usually ignored
by most manufacturers, since requires time, special equipment and test
methods. For sure that is ignored by all mass production manufacturers (Asia).
Other aspect in question - coatings. The math is not needed to show
difference in cost of coating of 2 outer surfaces of oiled objective vs. all
6 surfaces of airspased objective.
No math also needed to show the winner in light transmission; light loss on
reflections from 6 surfaces of airspaced design is ~3% in the middle of visual
range and 6% and more on the edges of it - and this is in case if we have
perfect coatings; if it is an average only - the light loss on the on 6 surfaces
on
the blue end of spectrum may exceed >10-20% (deep blue looking coatings),
that will shift color balance and make background yellowish. In the oiled
design the light loss on reflections is only 0.5-0.7 %.
What is next? - Scattering light, on six surfaces three times more than on two -
where is the higher contrast to be? How about accumulated errors on 6
surfaces vs. 2 - for sure 6 surfaces has to be done on more precision level to
reach quality of two outer surfaces of oiled design - that takes more time and
increases the cost of production. Will accumulated errors affect Strehl - why
not - in best case same 3-5% or even more of polychromatic Strehl will be lost.
BTW, the third order spherical aberration of 1/4 PV resulting dramatic Strehl
lowering to 80%! That is the area of attention - the optical surfaces must be
done on highest level possible!
You may see know why we choose and stay with an oiled design in all of our
APOs.
About glass choices, or in the other words, any advantages for CaF2.
Even first three APO160 F8 were done with CaF2, we had a plan to produce
APO160 with ED glass (price difference for blanks for APO140ED vs.
APO160ED is only $400) and be priced for appr. $8000, but we went to
Fluorite. Why? Our work experience and testing the ED glasses and CaF2
blanks in sizes up to 200mm and larger showed us that homogeneity of high
quality single crystals of CaF2 was better than in ED glasses, same time the
refractive index of CaF2 is constant up to 5x10E-6, that means no
measurements are needed; no refractive index variations, as in ED glass
blanks of sizes over 6" diameter. As being said before, CaF2 by being a
crystal has near zero internal light scatter and after all the Fluorite design
gave us possibility to force focal ratio to F7 without compromising color
correction in wide range including blue end - same level in practical
realization on the as in the best airspaced lenses.
Best regards, TEC
영어가 짧아서 뭔 말인지 도통 알 수가 없군요.
.
저도 과거에 왜 오일을 넣었는가 궁금했던 적이 있었습니다. 그 당시 제가 주장했던 것은 오일을 넣으면 (1)분해능이 올라간다 (2)수차가 줄어든다라는 것으로 주장했습니다. 처음에는 (1)번만 주장했었는데 말이 나와서 토론이 되다보니 다시 한번 이리저리 따져보았는데 분해능이 올라가더라도 아주 적은 수치라 생각되어서 이걸 보고 설계자가 오일을 넣을 리가 없다고 생각했지요(그 당시 박병우가 황당한? 주장을 던져놓고 꿰맞추려고 이리저리 자료를 뒤적이니 현미경에서는 분해능을 올리기 위해 오일을 넣는 현미경 자료를 발견했습니다. 이건 분해능 향상 목적이 아니라 다른 목적이라는 의견도 있었지만 논리적으로 현미경에서는 분해능 향상이 오는 것은 맞습니다.). 물론 제가 자료를 보고서 한 말은 아니었고 렌즈를 따지다 보니 분해능 이야기가 나왔습니다.
.
그러나 분해능은 제품 설계에 그대로 반영하기가 무리가 있는 것같아서 달리 생각해보다가 (2)번 쪽에 줄을 섰는데 그 당시는 뭔가 근거가 있어서 그랬지만 지금은 잊어버렸습니다.
.....................
.
그런데 준희씨가 올린 자료를 보니 위의 (1), (2)번과 영 다른 이야기같군요.
.
위의 글에서 용어가 낯선 것도 많고 영어 실력도 부족해서 요점을 파악하기가 어렵군요. ‘Refractive indexes’라는 말이 나오는데 이걸 ‘굴절계수?’라고 하는지 모르겠는데 이 용어는 본 적이 없습니다. 그리고 그 값도 아주 작은 값이군요.
.
아뭏던 준희씨가 올렸으므로 번역도 같이 해 주심이 좋은 줄 사료되옵니다. 눈 먼 백성들을 위해서요.
.
ps)
통상 모든 제품의 설계에서는 이론 목표치와 실제 목표치가 존재하게 되는데 그 적용 유무에 대해 가장 고민을 많이 한 사람은 제품 설계자입니다. 또 가장 광범위하게 검토하는 사람도 제품 설계자입니다. 위의 글은 제품을 생산하는 사람의 글이므로 현실성이 있는 내용이리라 생각합니다.
.
과거에 냉장고에서 직냉식이 좋니 간냉식이 좋니 한참 씨끄러웠던 적이 있었습니다. 위의 글 서두에 언급대로 두 방식의 장단점이 있습니다. 만약 한가지가 좋다면 제품 설계자가 두 가지 방식 설계를 할리 없지요. 허나 엘모전자는 간냉식을 생산하므로 이리저리 간냉식 좋은 점만 둘러대서 간냉식이 좋다고 할 수 밖에요. 현재 국내 주력 냉장고와 아시아존에서의 주력 냉장고는 모두 간냉식입니다. 토시바만 직냉식인데 대우전자가 토시바와 기술제휴를 하다보니 직냉식으로 가버렸는데, 이제 토시바는 가전에서는 삐리리로 전락한 회사가 되버렸는지라 직냉식은 소형모델에서나 사용하는 방식이 되버렸습니다. 허나 유럽에 가면 아직도 모두 직냉식입니다.