I recently purchased a used C14. A black-tube model, from about 2000. When I got it onto my G11 mount, I was having trouble building a model, and then realized that the image was shifting as much as 25 minutes of arc as it was pointing in different directions. Collimation was also going significantly out in different parts of the sky.
After reading many how-to threads, I removed the corrector, built a custom pin-wrench, and tightened the retaining ring on the front of the mirror. It is now down to about 4.5 minutes of image shift. Focus shift movement alone is about 0.4 minutes. Seeing has (of course) been lousy since I did this, so it's hard to get a really good collimation, but there still appears to be a small change associated with the flop (in part of the sky the secondary shadow looks centered, in others it seems to be a tiny bit off, but there is also heavy distortion with heat waves from the ground).
Is this amount of flop considered reasonable in a C14? Or should I try tightening it more? The ring is firmly hand-tight, done with the mirror face up, shipping locks disengaged, and focus varied between attempts to turn it. I don't want to pinch the mirror.
That's the basic question of the post.
Back story for those who are interested and to anticipate some possible questions:
The previous owner replaced the tube with carbon fiber, and in the process put a chip in the edge of the corrector (only about 3/16" by 3/8" of it extends beyond the ring and is blacked out). I have been in touch with Celestron, and they diagnosed it as mirror flop, and said to send it in. But with a chip in the glass, that seems more risky than normal, and there is no original packaging (the OTA was deforked). So I hope to avoid that.
In the process of getting to the diagnosis, I have checked the secondary to ensure that the collimation screws are all tight, the secondary is not loose on its plate, and the secondary housing isn't loose (it's an early Fastar model).
The measurement of flop is by carefully centering an object in the finder (I've used three different finders and mounts to ensure it's not finder flop), then comparing the position of the object in the eyepiece to the field of view. First an object near the meridian in the south is carefully centered in both, then the mount is flipped to the far side of the meridian, then the comparison is done. I flip back and forth at least 3 times to check repeatability. Centering the object in the eyepiece produces a clear displacement from the crosshairs in the finder, which eliminates a parallax effect with the hairs as a source.
Initially, an object that was centered on one side would be just beyond the 42 minute FOV of a 41mm Panoptic on the flip side. After tightening, it is about 2/3 to 3/4 of the way out from the center of the 11 minute FOV of a 10mm Delos.
I have also tried two visual backs and three diagonals to ensure that there isn't flexure or miscollimation in the tail end of the optical train, and I have tried flexing and wiggling the corrector to ensure that it isn't loose. Note that the method does maximize the potential for flop, as it puts the OTA almost horizontal on one side then almost horizontal on its opposite side.
What remains seems to be residual mirror flop, with perhaps 10% attributable to focus shift. There is a small change in focus with the flip, and the image shifts less than half a minute when reversing focus direction, which I assume as a component contributed to the total by the slider tipping on the baffle tube. I haven't pulled the mirror and slider assembly to regrease it. All of the tightening was done with the mirror still in the OTA.
I'd like to know if this is still out of spec for a typical non-HD C14, or if it's now in the normal behavior range, and I should either stop worrying or look to an alternative such as flop stoppers and a Crayford focuser, rather than torquing more on the retaining ring.
Thanks in advance for any guidance.