- This topic has 98 replies, 3 voices, and was last updated 7 years, 4 months ago by
.
-
Posts
-
Sincerely Massimo are not expert in star test and do not know be helpful…
However, it seems strange that you can not notice appreciable differences by increasing or decreasing the distance between primary and secondary…Although the ball was due to the mirror plane, Also like you said, you should still see an increase or subtraction of moving spherical mirrors…god!!!
You do not have the ability or know someone who has a flat mirror to make tests in Autocollimation?
Have you tried doing a test or Ronchi foucault directly to the focuser? maybe some clue might give it to you…And, It seems strange to me…
I tried with the Ronchi grating, the lines seem pretty straight, but my network has a high spacing 2 linee/mm, a failure of this magnitude ( 1/4 wave ) I do not think is discoverable, definitely it is not visually and in conditions of very little brightness, You should at least take a picture and examine increasing the contrast but at the moment I have no “accrocco” suitable for the purpose, the light is so little that the compact or webcam , even without lattice, They do not even realize that there is something to be photographed…In fact, I also pictures that I tried to do Ronchi on a star virtually unseen…
But with the Foucault knife placed right next to the fire I can notice many details that the Ronchi usually do not see…Or better, you can not see very well if the entire surface tends to become “gray” simultaneously or if there are bad areas or is spherical…seeng permitting of course…And, It is to try
However if it were spherical of that entity, it seems in accordance with the fringes of Newton, that indicate respect to the gauge a residual deformation of 1/4 The lambda, and in some places maybe something more …
I think I read the book of Texereau that the spherical you see in the Carregrain subtests, in some cases it is not decisive. I have to search the translation (to be reviewed) I had done Chapter).
And, if it were only a sub-correction of the secondary ( within limits ), it does not affect the goodness of the system that could be optimized in the distance between the mirrors as a function of the actual K of the secondary.
But in this case we can not know for sure, because the test detects a spherical star, although minor, the system as a whole, but it tells us nothing about what the cause
Reflecting well on this thing he said to Giulio, I think he's right ( for a change…
)I am seeking a better spigarmi, because it is important:
The gauge and the convex are constructed and tested for a conical constant K = -5.36
This non It means that the hyperbola is verified only and exclusively for that value of K.There is a tolerance zone which allows the same hyperbola to be verified for a certain range of K, depending on the precision of the figure.
Even, in my case, If the final test of foucault, with the same readings, imposed on the value of K = -5.38, hyperbole is correct to lambda / 15 !
Taking the evidence with the latest readings of Foucault, I have seen that the hyperbole of the convex is verified with accuracy of better than lambda / 4 in a range that goes from -5.30 until -5.43, a beautiful “scissors” !
What good is all this ? It serves to make sure that the RC system is optically not occurred for a total focal alone, but for a corresponding range of values, in function of a central ROC of mirrors.
( The central ROC is the core value to be known with precision and which can not and must not be changed in the calculations )
Then, moving away or approaching the mirrors, within the limits described, the focal resulting changes, but the configuration is verified and do not introduce aberrations. Of course there will be an intermediate value when the system is at full potential, and this can be calculated, but in the rest of the scissor permitted, the system is still correct.
This explains why when I tried to move the mirrors, I did not notice any additional spherical, simply because there could be, the displacement of 2 indeed mm, It allows my system to stay within the tolerance range of both primary and secondary.
But there is still a spherical, that seems to remain constant and at this point I think it might be due to a bad alignment and centering of the optics. Perhaps I underestimated the fact that a very small change of the secondary decentring the hyperbole making asymmetrical with respect to the optical axis and consequently varying the curvature at each point of reflection, much more than you can make a small variation of the conic constant.
So I think it's to check all alignment trying to do it with more precise methods of how I've done so far ( eye
)The Richtey-Chretien is also famous for this, collimation understood as compliance with all optical and mechanical axial squareness is awful and certainly I have faced so far this aspect too superficially.
The reasoning row and I agree fully with what you wrote….
Were you able to calculate how much you should dismiss or allow the mirrors to the ideal position to get an aberration of lambda / 2? o lamda / 4 o lamda, in short, the one you want, only to realize the orders of magnitude involved…
Then, this is the interval ( misure in mm ) in which , according to the previous discussion, the secondary verify the config. with not less than lambda correction / 4.
In the table there are no corresponding changes on the primary K, as they are of two lower orders and therefore irrelevant in this range. ( this is why it is said that: “is the secondary that must be built on the primary”, It is much more difficult to make a key that fits the secondary, obviously only in the RC )
focal ROC K Mirrors Distance Back-focus 2233 756.038 -5.42 518.82 203.7 2260 755.994 -5.295 517.2 218.5 beyond these values, my secondary continues to occur ( with ROC assigned 756 mm) the system and therefore greater / lesser distances between the mirrors but its correction for the corresponding values of K exits the tolerance zone, We should do the simulations, I have not done, to see which values of sub-over of the secondary correction is generated a spherical lambda / 4 on the system.
Interesting….
However, playing a little’ with ATMOS, using data from your optical configuration, I saw that (with ideal mirrors, then optically perfect) to introduce spherical aberration of lambda / 4 to the optical system, We must dismiss or allow the mirrors of about 3.5mm while the lambda / 2 aberration that amount rises to +- 6.5mm…
Which it is a great margin.So I am feeling thought that the tolerance in the distance between primary and secondary was moooolto more thrust, as I see it is not so. Also to get used to a bit 'of numbers, if you fail the installation of mirrors and mounts them 1 mm closer or further away, It is introduced into the optical system spherical aberration equal to lambda / 13.5. In short, quite small in relation to the value of 1 mm that already consider a wide margin and easily improvable also at amateur level or in any case non-professional.
I do not know yet instead evaluate, that could affect the final yield a decentring of the secondary.
Beautiful analysis, which roughly matches the one that I can evaluate in practice, although it seems to me that there is a margin slightly’ lower, but it is an evaluation made ” eye”..
Question, you did the calculations and maintaining focal K unchanged after moving the mirrors or you have calculated , in function of the displacement:
– the new optimum value for the conic constant .
– the new overall focal .The decentring think it can be calculated by assuming that:
one of the secondary lateral displacement is equivalent to “to shift” the radius of curvature of that point to another point. It would be like decentralize the mask couder, or the relative area centers on the calculation software while maintaining the readings.It also would add astigmatism, because the radii of curvature would no longer be symmetrical with respect to the optical axis.
Anyway, I believe that the values that you are getting the ball to be reported to the optical axis, it's correct ? because unfortunately things change very quickly for off-axis aberrations, this is also visible with the simulations of the commercials for a particular field of view.
Nerd, I kept focus and constant K.
I thought I'd see how the situation evolved, two mirrors data, to vary the distance between the two.
I'll show you the video of how I did, maybe I've made some nonsense without realizing.
However as you can see, varying the distance of the only coefficients of the polynomials Zerniche mirrors that change in any appreciable way are only those relating spherical aberration, while those of coma and astigmatism virtually no change.
Of course, with every change of the distance of the mirrors I must say performing this test in “best focus” because, obviously, It will move with respect to the position in an ideal position mirrors.
Of course it is right ! the spherical increased nearly nothing and you do not see a coma and astigmatism with a point source because we are exactly on the optical axis.
But we also need to understand what happens to the image in what has all the correct field ( or what purports to be ) project.
Try to increase the field… (half field angle – and direction)
- You must be logged in to reply to this topic.