astronomy9

The RC16 initial installation is described in equipment 4, but here its hardware is further developed.

I had wanted full frame through the primary mirror guiding, but this can only be done with a cold beam splitter (dichroic), that reflect visible light at 90 degrees toward the imaging camera, and the remaining infrared light toward the guide camera. I contacted various cold mirror manufacturers and settled on Alluxa who sent me a large format 1/12th wave overcoated cold mirror that cuts the light at 695 nm, allowing the main imaging camera to receive SII, OIII and Ha emissions.

To house the mirror, I designed by OpenScad a 3D printed housing with a slot to accomodate a 3D printed mirror holder. I painted the housing, and machined an aluminum box with three ports and threads so it can be mounted onto the RC16 and take on the imaging and guide cameras. I also included a 0.5X reducer for the guide camera so that a CMOS sensor half the size of the imaging camera would lead to nearly identical image fields on both cameras. The image below shows the cold mirror housing, the guide camera, and the GSO stock focuser modified with a linear actuator.

The very first testing under star light was a revelation - no other moment in my 45 years of Astronomy has been so, save for the first view of Saturn or lunar craters. The 2 second exposure at the guide camera in the infrared was teaming with stars, regardless where I pointed it. Guiding will be a snap, despite the very weak ghost image that occurs from passing light through a beam splitter.

These images show the modified GSO focuser, the dichroic housing, ASI6200MM PRO imaging and ASI294MM guide cameras. Of note, the 0.0125mm per step linear encoder is seen on the left side of the optical train, and the linear actuator on the right side. A linear actuator coupled to a linear bearing (mono-rail) focuser is an excellent replacement for the shaft on rail friction which slip at these imaging train weights.

To close up the OTA from dust and insects when not imaging, I ordered a cloth shroud from AstroZap and used OpenScad to design large paddles, 4 of which actuated by 270 degree RC servos will close up the OTA's aperture. When printed, these paddles are only 67 grams each. The mounts for the servos is not shown.

The linear actuator support body and clamp were most easily designed when for free I downloaded the linear actuator mockup! OpenScad easily lets one fit the body to the focuser curve and add the proper holes. The linear actuator is best driven by a PWM motor controller which the Pololu Qik 2s9v1 Dual Serial Motor Controller did quite well. It does require a serial TTL stream, which a USB FTDI board did in a snap. PWM improves low speed control, and despite the 250:1 gearing of the actuator, it needed the PWM board to move slower.

And... months went by, the AstroZap shroud wasn't coming (Covid...), cancelled the order, bought a Singer 4411 sewing machine, some yards of cotton, and sewed me a shroud, all for about the same cost as the AstroZap. This allowed continuing the design and fabrication of the main aperture shutters and writing the code to drive a Pololu Maestro micro servo controller.

Telescopes aren't ready out of the box for automation. For one, an RC is full of orifices that dust and insects infiltrate and compromise the optics and optical path when not in use. Hence the sewing of the shroud, the numerous 3D printed parts - by way of an example, I traced and scanned the outline of the middle truss plate 'holes' and used various PC software tools to detect the outline and import it to OpenScad as a set of points, which was then easily turned into plugs (normal and mirrored) with tabs that nicely fit into the 'holes'. 3D printers are a blessing - 79 parts all by OpenScad, some 3 kgs of PLA, are part of the telescope alone. Some parts, such as the shutter paddles, were modified and reprinted numerous times as I adjusted and tested the design, these even hinted at a problem with the 3D printer which I then corrected (X & Y axes were not perpendicular), so the total parts printed are well over 100. You can find some on Thingiverse.

Above are shots of M27 from both imagers to illustrate the differences when star light is split at 700 nm - unsurprisingly the emission nebula is not visible in the guide image and indeed the stars look different. 30 sec on the main imager, 10 sec on the guider. Both camera frames are shown at the same angular scale for comparison - with the guide image frame size shown on the main imager as a gray outline. The main imager is 36x24mm whereas the guider is 19.1x13mm - but the reducer (158mm RFL) is not achieving the 0.5x because I could not place the reducer at its specified working distance of 79mm, but at 40mm which yields (RFL-WD)/RFL=(158-40)/158=0.75x, and so the inverse of this is used to scale up the guider imager size for comparison, or 19.1x13mm divided by 0.75 gives 25x17mm, which leads to the guide image appearing 25/36 smaller or 70% as large as the main image. Of course it could have been calculated in angular extent to achieve the same result. I chose the ASI294MM because its back illiminated sensor was particular good in the IR, 14 bit default and binned at 3, the sensitivity is remarkable even at 2 sec exposures - a very pricey guider, but it can also serve as a high speed planetary camera.