Astronomy - Andre Germain

The beginnings of a lifelong interest in Astronomy was kindled in 1979 when I chanced upon books in the town's library. I remember well the photographs of primitive but effective telescopes, noisy galaxy images, and spring air at the time. Decades later, the same books are no more at the Westmount library, but the same spring scent evokes the memories.

Astronomy was a private affair until the mid 80s when I visited the folks at Stellar Telescopes on Decarie in Montreal; Fred Clarke, Donald Alexander, Gary Boyle and Alister Ling. Friendship ensued and we'd meet Thursday evening at Don's and late summer at Stellafane. Fred passed away in '98. The CAFTA astronomy club named a yearly prize after him for his selfless help to members. My astronomy is private again after Fred. Donald also departed our world in '06. Two stars in the firmament shine brighter.

Having graduated in both Physics and electrical engineering at the university of McGill in Montreal, my interest in Astronomy covers the latest cosmological developments to telescopes nuts and bolts. I've polished the 8 inch mirror in Cassiopeia (silly name) twice to better than 1/6th wave, and built three mounts in the process, coded thousands of lines of C, built an observatory and two CCD cameras. I've ported snippets of my code onto flight simulators! The mostly home made components are discussed below; I've spent too many hours on glorious summer days and freezing winter nights coding outside, hammering and cursing my designs, but I don't regret it at all in this day and age of low cost off the shelf. It's near impossible to be a renaissance man/woman anymore.

Dome

Having built a house in 1995 on two heavily forested acres, I erected a tower to catch a better view of the firmament rolling above. The concrete pier sprung from deep in the ground and isolated from the octogonal tower is some 18 feet in height. The observatory stands 22 feet and 12 in diameter. Four home made drives move the dome. The unfortunately narrow 29" dome slit is opened by chain driven top and bottom screws, powered by a car battery charger through contact points at the dome park position. Relative position and velocity are sensed by an optical encoder and absolute position by an AS8150 barcode reader with many barcode strips along the inner periphery of the dome. A pan/tilt infrared night vision camera (modified webcam) provides a view into the dome during remote operations. See time lapse videos here.

Telescope

The Newtonian 8 inch parabolic mirror has a 44" focal length (F5.5) and hand figured twice since 1980. It tied for best optics at Stellafane in 1989 (the other was a Maksutov 6"). The mirror is the only original part of the telescope; three paper and one aluminum tubes, three mountings, two secondary mirrors, many new coatings later, it still is my workhorse. A 90mm Maksutov on a Goto mount (iOptron's "the Cube") serves as steerable guide scope via Ascom - arguably the only such use of a Goto scope in the world - and works well to find a guide star! I no longer use these scopes for visual work as I do all my work remotely from the house via wireless and cooled CCD cameras. Both scopes have full aperture RC servo motored sky end caps; stops the dreaded Asian Beetle from crawling in.

Mounting
The fork arms were cut and welded to my specification. I milled and lathed parts, including both nylon gears to mesh with 3/16" bolts; this was done by spinning a tap against a nylon disks on a lathe, it naturally cut into and spun the disk. I built and coded an 8051 uController four channel power amplifier to drive the phases of two stepper motors (declination and ascension axes) in a saw tooth, that's right, not micro-stepping! It's ultra smooth, and is advanced by receiving pulses from an ISA card in a PC under MS-DOS. Unfortunately, it does cause more stepper motor heating, and any heat plume is undesirable near optics. Right ascension screw sensing provides PEC (periodic error correction) with either manual or automatic guiding. The same DOS PC also has predictor corrector tracking software to ramp up the motors to track satellites via LX200 RS232 commands issued by satellite tracking software. The sky charting Guide 8 software provides pointing commands via the network.

In July 2007, the stepper motors and home grown linear amplifiers were replaced with IMS intelligent motors (capable of 10 deg/sec) linked via USB, and the incremental encoders were replaced with 17 bit serial absolute encoders that are read by a USB FPGA (see here) with verilog code I wrote. It had become a pain to support both the ISA card and the relative encoders which required finding the index pulses for absolute positioning. The DOS PC was no longer required and removed from the observatory. I coded Ascom compliancy into my drive software so the CCD camera could guide the scope; Ascom instead of RS232 since both are controlled by the same PC. In November 2007, the right ascension gear and worm were replaced with a 9 inch Byers gear and worm with a far lower and repeatable periodic error (shown in graph below - deviation of star spot in 1 second exposures at 10 second intervals).

CCD Cameras

I built two CCD cameras; the Cookbook Kodak TC245 followed by the Audine Kodak KAF-401E. Both are cooled with Melcor's Peltier thermo-electric ceramic modules and pumped glycol; this is necessary to improve the signal to noise ratio of the CCD chips. The Peltier current is controlled by PWM. The Cookbook was a heavily modified version of Kanto, Berry and Munger's design, with 16 bits conversions and reams of home brewed code and was subsequently replaced by a USB2 Starshoot by Orion which guides the scope via Ascom. The Audine, designed by the French Aude group, serves as imaging camera, and has a novel balanced impulse shutter I designed which produces no heat in operation, consuming only energy when a change in position is required. On my travels to Tenerife, I had an Audine USB box shipped from France's Essentielles Electroniques which has greatly improved the camera, ridding me of PC resource hogging that was via the paralell port, halving readout times and removed the interference picked up from the PWM temperature control that plagued my images. It has completely rejuvenated the camera. The Audine is known for having frost build up on the CCD glass face, so foam blocks were sculpted to fit in most of the empty spaces within the camera (shown below); these are an ongoing improvements, and will require moving to pumping a partial vacuum (on hand, awaiting installation). The coolant is in a tank, circulated by an R/C hobby fuel pump, with valves to help purge or replenish the system easily (see photo further below - click to expand).

Computers and Software

Back in '95, an 8086 first drove the dome and scope. I had designed and built an ISA bus card (visible in open box image here) with 35 ICs to provide all manner of digital and analog I/O, incremental encoder counters, pulse output and such. It also provided reading in 8 bit wide data from the Cookbook iso of the paltry 4 bits. Buckets of code written in C to interface with the dome, scope and CCD via this card and make it all work like it should. All migrated to a 80166, 80286, and finally a DX400. At that point, a Pentium II joined it to run Guide and the new addition of the Audine. Lately the Audine has moved to an Athlon 1.5 GHz that supports USB2. The DX400 and Athlon link up via coax ethernet to another PC in the house (upgraded to wireless in 2007), where I do all my control remotely, a godsend during Canadian winter and mosquito infested summer nights, to say nothing of the Asian beetles. At least I got smart with the Audine and used freely available s/w to operate it, but I still had to build my own PWM Peltier cooler with an 68HC811 via RS232. I coded lots of s/w to reduce the data, mostly batch processing of the raw images with their dark and flat field frames, and a clever sub-pixel alignment tool which pixel interpolates multiple FIT images to render improved resolution and star shapes, somewhat like NASA and Meade's drizzle. PC hard disks life span are much shortened by freezing temperatures, so both PCs are inside an insulated box (shown here open and closed - click image to 'open'), with three thermostats; one controls a low power hair dryer to increase the box' temperature, at which point a second starts up the PCs and a third exhausts excess heat via a fan. It is also fully sealed/screened from insect intrusion. Diskless operation with CF flash memory via IDE was attempted to remove the need for temperature control, but proved to be excessively slow and having too few total permissible write cycles.

A recent addition is a Pentium D, unheated, which controls the dome and slit, pan/tilt camera, Starshoot CCD and other USB devices. An Hitachi Endurastar 2.5" hard disk was chosen as it can operate down to -20 Celcius. In July 2007, the home grown drives and incremental encoders were replaced with commercial components accessible via USB, and so the DOS PC was no longer required. In May 2008, the *warm box* PC was removed and all operations and devices are now under the lone Pentium D PC. Under XP and with remote desktop, full coding and testing remotely is the norm - most welcomed in winter.

With all the money spent over the years, a large Meade and CCD could have graced the dome, but it would not have been anywhere near as much fun in the process!

Pardon the mess of wires in the image below: during improvements in the dead of winter, wires are very stiff. I'll clean it all up on a warm summer's day (...I finally did!).