Often the bad images the telescope primary mirror is not responsible or bad collimation optics, but just plane mirror. Plane mirror should have a precision of lambda / 4 and if it is issued with one, This will produce very good images. The best form of checking the quality of the implementation of a flat mirror is interferometer study. Unfortunately this is expensive equipment and thus less accessible to astroamatorów. It is much easier with a glass test interference. But while the study by the interferometer is a noninvasive, in the case study of the mirror glass test must be applied directly to said test surface of a mirror. And here arises the problem, because to do it right, You must have some experience. Firstly, both the surface of the mirror under test and the test must be absolutely clean. Any contamination between the mirror and the test will result in inability to properly measure, the additive risk of scratching the mirror test and. I would add, that while the individual features of the primary mirror are virtually no effect on image quality, is very scratched plane mirror can unfortunately worsen given images. Using scratched test increases the risk of damage to the surface of the respondents, harder it is also properly applied the test. Some people examining the surface using an interference test, put a mirror between the test strips of thin paper, and thus do not permit to cross the surface damage (This is obviously only possible when testing flat surfaces).
Application of test to the test surface with three dividers paper. Only possible when checking flat surfaces
This method has one major drawback. Well, generally obtains an image of a large number of thin bands. Of their shape harder to really authoritatively judge the quality of the surface. The ideal is to get a picture, say, three bands, one passing through the center of the mirror and the two sides at the edges. It is worth examining the two sides of the mirror, so that for the first time bands moose ran along the long and the second time along the short. The following image obtained during the test surface using an interferometer made by me.
Image very well done flat surface, interferometer test my design..