When Astro-Physics announced their new .67x focal reducer, some quick calculations indicated a possibility of achieving the required working distance with an AO-7 to achieve this reduction ratio. Brad Ehrhorn and I cooked up a design to attempt to meet this requirement.
The design consists of remounting the AP optic in a new holder that mated with the 2.7" threaded Precision Instrument Rotator. The other end of the holder mated with a modified AO7 diagonal. The intent was to get the reducer lens as close to the mirror as possible.
Copyright © 2003 Brad Ehrhorn and John Smith
The upper image shows the completed assembly. The lower image shows the pieces and parts. Note that, unlike Russ Croman's Takahashi reducer assembly, there is no provision for rotating the camera - this is a PIR solution only currently. Of course, it should be considerably less expensive. To remove the reducer for native operation, there are two solutions. One can either replace the entire assembly, modified diagonal and holder, or remove (carefully!) the optic. The resultant aperture is larger than the standard AO7 and should give a bit more vignetting immunity. Interestingly enough, removing the optic and adding a 1.75" spacer achieves focus within .24" of back focus distance. In my case, I was also within .2" of correct back focus with the camera and reducer mounted directly to the focuser. Since the 12.5" RC design has the least amount of back focus distance, this is good news.
Initial testing with image link indicated a reduction ratio of .66x - pretty close by my book. With my 12.5" RC at F/9 and an ST-10XME, my image scale was 0.49 arc-sec./pixel. With the reducer, it was 0.74 arc-sec./pixel. Although the star sizes were quite similar in FWHM arc-sec. with and without the reducer, some collimation errors were noted on the right side of the image. I suspect the reducer makes these errors more visible than the native F/9 operation. Seeing was on the order of 2 arc-sec. or so, and the resultant star sizes were definitely comparable. More testing under better seeing conditions would be desirable.
I took a quick image of the Cocoon Nebula, IC5146, nearly full moon and all, to get a feel for the resultant FOV.
The upper image represented around 4 hours of data taken at F/9. The lower image was around 2 hours of data taken at F/5.9 with the reducer. The increased FOV provides for a more pleasing presentation, at least in this case.
I plan to do more work with the reducer, now that the proof of concept is complete. I certainly need to touch up my collimation.
Thanks to Brad for working with me on this and building the prototype.