Brutus Modifications Round 3-- "The final chapter"

...a six month saga!!!

Customized "On Demand" tracking system

This was the heart and soul of this round of modification to the scope, and the one that ended up consuming an embarrassingly long time to implement (don't even ask how many hours it took in total-- suffice it to say that there were 3-4 unsuccessful builds of both the altitude and azimuth systems before the final version you see illustrated here!).   My goal was to have "on demand" high accuracy tracking without disturbing the silky smooth motion of the scope when used manually.  I am 'old fashioned' in my taste on Dobs, typically eschewing  Digital Setting Circles and having no wish for a "Go To" StellerCat^tm type of automation.  Rather, I anticipate continuing to use the scope manually probably 99.9% of the time, and want to reserve tracking primarily for high powered planetary observing or public outreach sessions.  In short, I didn't want to disturb the marvelous ease of using the scope for a mode that I would use only a handful of times per year.

The core of my automation was to be the Dob Driver 2 (DDR2) system produced by Tech 2000 (http://homepages.accnorwalk.com/tddi/tech2000/driverdescrip/).  I recycled the DDR2 unit previously installed on Frankenscope and upgraded the motors from standard speed to a set of Ultra High Resolution (4X as much torque, and motions 4X as fine) acquired gratis from a fellow NOVAC member.  The 'issue' with a standard Dob Driver installation is that it requires changing the 'feel' of the scope -- in the case of the azimuth axis by removing the Teflon pads and putting in steel rollers; in the case of the altitude axis on a scope the size of Brutus, by putting in rollers and possibly an aluminum facing on the altitude bearing surface.  I had changed the azimuth motion of Frankenscope as required-- and in fact went beyond the level specified, adding a 'lazy susan' as well as the rollers-- but found that a system with mechanical bearings and rollers is hard pressed to be as smooth and trouble free as one based on the friction between two smooth surfaces (Teflon and Ebony star).  My solution would be to change axis of motion of the azimuth motor from the VERTICAL direction-- where the weight of the scope produced the requisite friction for successful operation-- to the HORIZONTAL, where the drive wheel is 'pulled' into the side of the circular ground board.

AZIMUTH Motion:  I therefore made a design decision to deviate from Tech 2000 orthodoxy and mount the knurled drive wheel horizontally instead of vertically.  This horizontal mounting requires making the knurled wheel which turns the scope's rocker box bear against the outer edge of the ground board, instead of vertically sandwiched between the rocker box and ground board.  I had seen modifications of this nature that used a pivot and opposing springs to pull the wheel into contact with the ground board edge, but it seemed to me that this "wasted" most of the force being generated by directing it tangentially away from the center of the scope.  I designed an approach that vectored the force of the spring radially inwards almost directly toward the center of the scope.  I decided to use a bicycle brake cable to transmit the force, and it was long enough that I was able to run it around to the side of the scope.  This meant that the azimuth control would be located mere inches away from the altitude one. (On Frankenscope I had to walk around to the front and the far side of the scope to engage the gears.) 

   In the image to the left, notice that the short length of dowel the cable is routed around near the motor directs the vector of force inwards towards the center of the scope, rather than off at an angle.  Also see the section of truss tube (at left) used to route the cable 'gently' around the 90 degree angle of the corner of the rocker box.

Four rounds of modifications and a lot of work proved necessary, not so much in the mechanism of the pivot and wheel (which I mounted to a length of angle iron, and used a section of Oak rail as the vertical axis) itself as in its interaction with the ground board and with the control/clutch mechanism.  My ground board was not perfectly round, showing lots of minor dimples around its perimeter.  I faced the ground board's edge with aluminum flashing (Tech 2000 indicates that this gives ~40% better traction than unadorned wood), but my first attempt failed when the knurled wheel was unable to climb out of the 'valleys' posed by the depth of the dimples.  I stripped the flashing off and filled in the uneven surface of the ground board with wood putty to the extent possible.   I then put a new layer of flashing on.  This time the scope would turn, but the flashing was thin enough that the underlying residual unevenness of the surface would cause the wheel to stall out at unpredictable intervals.  Although I have been a firm believer in using powered graphite rather than the orthodox car waxes etc. at bearing lubricant, I broke down and tried silicone car wax.  It failed to reduce the friction enough to make a difference, and introduced "stiction" to the scope's fine azimuth slewing that took me a lot of work to eliminate.  A second layer of flashing atop the first ultimately solved the problem of inadequate "grip" nicely. 

The azimuth motor mounted (top and bottom view)      Rear view of the completed mechanism.

I had decided that a locking clamp from that wonderful purveyor of hardware http://www.mcmaster.com/ would be the actuating mechanism.  Since these clamps were adjustable and could be used to lengthen/shorten the cable by rotating the clamp element, I needed a way to be able to turn the cable in place without twisting it.  The solution I came up with was to use a heavy duty (#1 size) fishing leader swivel.  It worked nicely-- other than the fact that the wire of the first fishing swivel unraveled under the mechanical tension generated by my various trials.  (Its replacement is under less tension, and the swivel itself is coated with JB Weld epoxy, so I am optimistic it will not meet a similar fate.).   My early trials revealed that a single 20 pound spring could not generate enough force to pull the knurled wheel tightly enough against the ground board, so I ended up using two such springs in parallel.

                       The actuating mechanism and springs that generate the necessary tension to move the scope.

ALTITUDE Motion  In total, I spent ~50 hours spread over a month or so getting the azimuth motion right.  Once that was operational, I turned to the altitude axis.  The altitude dimension I expected to be trivial, since it was a conventional mounting approach complicated only by the fact that Tech 2000 did not design their belt drive hardware for use on a scope of this size and therefore some of the mounting hardware would have to intrude onto the arc of the altitude trunion rather than be mounted to the side of the rocker box.  I made a mount out of bar steel that allowed one of the two points of the azimuth pulley motor to be mounted suspended above the trunion, which solved this problem rather nicely (see below).

  To the left is a view showing my customized altitude mounting mechanism.  In order for there to be enough distance for the lower clamp mechanism (seen almost touching the scope handle at lower right) to be mounted, the uppermost of the two posts that hold the motor plate against the side of the scope had to mounted high enough that it was in the free space between the rocker box and the trunion. 

Once I solved the mounting problems of the motor and associated hardware, I installed the 11" wooden altitude wheel provided by Tech 2000 and fired the motors up.  I was confident that, since these motors were high torque and able to lift ~40 pounds, they would easily be able to turn the scope in altitude.  No such luck!  The problems were inertia --even though it is finely balanced, there is a lot of weight (~175 lbs) being suspended between these trunions-- and insufficient leverage (with an 11" diameter circle, one is trying to move this weight from close to its center of mass, rather than from its edge.)  One solution would be to further reduce the friction between the mirror box and rocker box.  According to the Kriege & Berry book, the scope's altitude Teflon pads are less than half the optimal size in terms of surface area loading.   I was able to get a fellow NOVAC member (thanks, Pete!) to give me some virgin Teflon so I could make larger altitude bearing pads.   I made the new pads 2 1/2 times larger than the originals, and planned to cut them down until I reached the desired point of tension.  The initial set turned out to make the altitude action HARDER than ever-- clearly I'd miscalculated the "sweet spot" and overshot back into the zone of too much friction.  I ended up using a combination-- two of the old and two of the new pads.  I faced the 11" Dob Driver altitude wheel with non-skid tape, and was able to create enough friction that the belt would bind and then "jump" (which is hard on the motor bearings), but the stock hardware was simply unable to deal with the ~175+ pound mass of the rocker/mirror/cage/tubes.  Time to get creative to maximize leverage, since even the highest torque motors couldn't overcome the inertia of this big scope.

Although there were exotic solutions involving multiple pulleys and gear wheels, the simplest solution seemed to be to make the altitude wheel itself larger.  The largest sized altitude wheel what would fit within the truss tube clamps on the mirror box was 23 1/2" in diameter.  (Even this required one truss tube clamps to have its tightening knob modified for greater stand-off, to avoid fouling the altitude belt.)  The standard Tech 2000 solution is to have a groove in the wheel to hold the belt in place.  Since I didn't envision having the wheel turn very far during any single session -- remember, even 90 degrees of arc would be 6 hours of tracking-- I opted only to put a raised ridge on the back side to make it easier to slip the belt on and off the wheel.  I left the wheel's circumference rough --only lightly sanded to remove splinters-- to leave ample friction.  This system obviously would not accommodate even the largest of the drive belts sold by Tech 2000.  Fortunately, these are standard 'micropitch' belts-- and McMaster Carr again came to the rescue.   I had written an Excel spreadsheet to calculate the belt length needed for any given diameter wheel, and then fiddled with the wheel diameter to get to a belt length that matched an available size.

  The oversized altitude wheel     The modification needed to enable one truss tube clamp knob's shaft to clear the wheel

Left: The micropitch belt at rest (draped over a truss tube knob)  Right: stored along the side of the rocker box

The system works like a charm, and may see more use than originally intended, now that I will probably use Brutus for a lot more public events in 2007. (see SNP Outreach)  The only minor 'glitch' is that I can't set the backlash on the altitude axis -- which may be due to my DDR2 control unit, which is old and starting to have problems with its electrical contacts.

Clement Focuser

Since I had sent my chronically malfunctioning 2" Clement focuser (http://www.clementfocuser.com/) back to its maker for overhaul, when it came back I decided to install it on Brutus rather than return it to the seldom-used Natasha.  It can accommodate a much heavier payload (and my large Naglers in the Paracorr weigh nearly 5 lbs), and with 3 full inches of focus travel might allow me to generate enough spare in-travel to use my favorite planetary observing combination (Nagler 3-6mm zoom and Sirius Variable Filter System-- VFS) unbarlowed at high power.  Part of generating enough in-travel was to shorten the effective length of the truss tubes.  Rather than cut them the needed 1" shorter, I mounted them lower in the wooden truss tube clamps-- actually, putting strips of aluminum bar across the bottom edge of the clamp block to allow the tube to seat all the way in.  This was a neat-- and reversible-- fix to the problem.  And yes, I generated enough in-travel to successfully use my Nagler zoom and VFS unbarlowed!

Clement focuser racked out                    ... and all the way in       An aluminum strip used to shorten the truss tubes

Trouble in paradise!  Once I had the Clement installed, I began using it.  The images were OK, but I could not pick out the "E" and "F" stars in the Trapezium at the heart of M42-- which I can usually do when the winter seeing is average or better.  I chalked this up to mediocre seeing, until a night of exceptional conditions when I still couldn't make these stars out!  Clearly something was amiss.  Daytime testing revealed that, despite its rebuild by the manufacturer, the Clement focuser still had the image shift that had plagued it since its purchase.  Since Brutus has a longer focal length than Natasha, the problem was in fact worse here.  When the focuser was racked from near its out-travel limit to near its in-travel limit, the aim point would shift across the primary mirror by a full inch!!!!  (see Developments and discoveries for images of this).  This would cause bogus results at every step of the collimation chain from laser setting the secondary mirror's aim point to the primary all the way through to autocollimating the tilt of the focuser relative to the secondary mirror.  I reinstalled the old Feathertouch focuser that previous owner Mark Dearing had put on the scope... and the performance problems went away!

Feathertouch focuser-- old and new

The best way to use an autocollimator is in concert with a focuser that has an adjustable base.  My elderly Feathertouch did not have such adjustability, so I tapped three leveling screws (one at the left center edge and one each on the upper and lower right edges of the mounting base) and added this capability.  This allows me to 'tweak' the collimation of the scope to virtual perfection.  However, the focuser draw tube has an excessive amount of play in it, which translates into sag off the optical center line of the focuser when a heavy eyepiece was installed.  I called http://www.starlightinstruments.com/, makers of the Feathertouch and learned that my unit was their oldest model, which was prone to mechanical wear.  I will send it back for refurbishment, but in the meantime the delightful proprietor of Starlight (Werner) will sell me a 'blemished' new and improved model at a discount. The new unit will have a 2.5" effective draw tube travel (vice 2" of the older one), so if I use a low profile 1.25 to 2" adapter on my Sirius VFS and return the truss tubes to their lowered position, I might still be able to get the Nagler zoom/Sirius VFS combination to come to focus unbarlowed.  (If not, I may cut 1/2" or less off of the truss tubes as a last resort...)

 I have been very happy with even my antiquated and malfunctioning Feathertouch, so the new unit should be nirvana.  (The old unit, once overhauled, will either end up on Natasha or remain mothballed until I get around to building my 10" travel scope.)  

The old Feathertouch mounting base with two of the three leveling screws visible on the base.

The new Feathertouch focuser has the dual screw brass non-marring clamp ring, a 2.5" draw tube, leveling base, and internal braking system.  This unit-- a discounted cosmetically 'blemished' (and discount priced) one-- has a 1/2" longer draw tube than my old Feathertouch, which allowed me to try to accommodate the nearly 3" of focus travel I need for the full range of my favorite observing toys.  By lowering the truss tube mounts back to the bottom of their mounting blocks (as I did for the Clement) and using a low profile 1.25" adapter, I could *almost* bring my Sirius Variable Filter System (VFS)/Nagler 3-6mm zoom planetary observing combo to focus without a barlow.  (Using even a 1.5x Barlow means the starting magnification is 600X+, which is far too high...)  Since I still couldn't quite reach focus, with some trepidation I then took the irrevocable step of shortening each truss tubes by 3/8" each-- which brought the Sirius VFS/Nagler zoom to focus with full quarter inch to spare!  While most of the regular eyepieces are racked far out, by my calculation most still have a 100% illuminated field.  Thus, I have regained the range of useable focus travel I lost when I removed the Clement focuser.

Update:  further calculation showed that, while all of the eyepieces would come to focus, those with lots of out-travel had a miniscule --or non-existent-- 100% illuminated area.  Even the big Naglers in the Paracorr have ~.3" 100% illuminated, which is pretty small.  (Berry & Kriege recommend between 1/2 and 3/4" 100% field for such eyepieces.)  It makes me wonder whether the scope before modification had any 100% field!  Moving the light cone another inch out gives all of the eyepieces a full 1/4" more 100% illuminated diameter, while still avoiding vignetting.  The challenge was doing this without cutting the tubes, in case I want to reverse course (or discover I've miscalculated.)  My fix was to make wooden 'elbows' that could be mounted under the truss tube clamps to allow the tubes to seat lower.  Thus, I gained the increase in fully illuminated field, while not doing anything I couldn't undo later. <g>

UPDATE: glad I made these mods reversible, since I seem to have reversed the sign in my formula.  I was actually decreasing fully illuminated field by the mod below, not increasing it!  I have reverted to the situation that *just* accomodates focus on the unbarlowed Sirius VFS/Nagler zoom, while preserving the maximum fully illuminated FOV for the mainstay eyepieces.  For the Naglers, this translates in to a 0.5" 100% FOV-- smaller than the 0.8 or 0.9" that Pat Rochford had when Brutus was built, but understandable given that I have created an extra 1.5"+ of in-travel (and the light cone shrinks 0.25" for every inch along the light path.)

To the left is a tube 'elbow' extension without the tube inserted, to the right is one with the tube in place

I use the leveling screws on the Feathertouch's base to true the focuser to the diagonal with the autocollimator.  In the default position, two of the leveling screws are blocked by the focuser knobs.  By rotating the mounting position of the focuser 45 degrees in the base, I uncover one of these leveling screws-- and end up putting the focuser knobs in a position that is more nearly horizontal under most observing conditions.

MINOR MODS:

Another handle for moving scope

When I was dealing with the inadvertent introduction of excessive stiction in the scope's azimuth motion, I installed an additional furniture knob handle on the secondary cage to allow for greater leverage and fine motion control in moving the scope.  Once I had wrung the excessive stiction out, I opted to leave the knob on, since I liked the extra options and flexibility it provided for fine motion control.

On-board 20 Amp-hour battery

Again sparked by the excessive stiction, I decided to replace the heavy (45+ pound) 85 amp hour marine battery installed on the front of Brutus with a lighter battery.  85 amp hours is overkill anyway, representing more power than I would consume in a full week of all-night observing sessions.  A 20 amp hour wheelchair battery weighs in at ~13 pounds, and still provides enough current for a full night of observing even with all possible electrical gadgets running. 

The marine battery (at left) and the wheelchair battery (at right).  I still bring the marine battery 'just in case', and to power the pre-observing front mounted fans.

Potentiometer control for the side-mounted fans that disrupt 'boundary layer' thermal inversion

This was one of those things that I had decided was optional.  However, at least one veteran observer at WSP had detected optical interference from the vibration of the fans running at full throttle (granted, this was at 820x magnification!).  As I had already begun to cannibalize Frankenscope for parts, it was trivial to take one of his two potentiometers and make the fans on Brutus variable speed.

The knob of the potentiometer is visible at left, hanging over the edge of the rocker box.

Labeling Kendrick dew heater cables

An absolutely trivial issue, but there are three sets of Kendrick cables running up a truss tube to the secondary cage, and I didn't know which was which.  Even granting that I have yet to use my Kendrick system on Brutus, it made sense to trot it out and identify which bottom end connector powered which top end Kendrick heater strip.

Fixing the DSC azimuth inaccuracy

This was an issue that had driven me crazy ever since the Winter Star Party in '06.  I eventually discovered that the small set screw/bolt that keeps the ground board from rotating relative to the main azimuth bolt of the scope was loose, and allowed ~30 degrees of slack motion.  It is fixed now, and the azimuth setting circle can rotate a full 360 degrees in either direction with sub-degree accuracy.  I thought about swapping out the 4000 tic encoders on Brutus for the 8000 tic pair on Natasha, but opted not to.  Higher resolution is not always better on a big scope (it might mean I'd have to rotate the scope more slowly so as not to overwhelm the encoders), and the cables with Natasha are too short to fit on Brutus in any event.  

'Star Party' observing tray holder

While I almost invariably leave the handles on the scope, conditions at star parties tend to be crowded, and the handles become a hazard.  Removing them means losing my on-board tray pole, so I made this replacement out of a 3 foot length of 2 x 4.  Not pretty, but it allows me to mount the tray even when the handles are removed.  I'll give this its baptism at the Winter Star Party. 

Extension tray for Sky Commander DSC control

Maybe my eyes are getting old, or maybe it's the fact that this tray table was made for a smaller scope (Natasha), but I find it hard to read the LCD on the Sky Commander when I'm at the eyepiece even when the Sky Commander is on the near side of the table.  I therefore made a ~8" extension that I can inelegantly clamp into place on those rare occasions when I set up the DSC's.  This 'plank' brings the Sky Commander box close enough to read easily.

        THE BOTTOM LINE:  I NOW HAVE A 24" DOB THAT DOES EVERYTHING I COULD CONCEIVABLY WANT.  I'VE RUN OUT OF THINGS TO BUILD AND TWEAK ON THIS SCOPE!  

Time to move on to another ATM project. ;-)