Iterations R Us
You may recall (and if you don’t there is a convenient blog entry you can scroll down to refresh your memory – I had to) that when we ordered the additional planetary gear for the azimuth drive, the planned integration would have modified an L-bracket holding a flange bearing so that it would now carry the output end of the planetary gear. The output shaft would then slip into the Trantorque on one side of the drive wheel, and the extension shaft on the other side of the wheel would almost meet it inside the wheel, in the grip of the other Trantorque.
I was concerned about that gap. But I was willing to try it out, in the name of keeping the drive in relatively the same form. Larry took one look at it at the last board meeting and said, “No.” That was all it took to move on (I nursed my shattered ego privately). With a couple of ideas batted back and forth, we settled on extending the assembly to accommodate the planetary gear, without much disturbance of the existing extension shaft holding the wheel between two flange bearings.
This iteration required a pair of extension blocks with a new mounting plate for the SiTech gearbox. We re-used the holes where the SiTech gearbox had been mounted, drilling them out to accept four 1/4-20 bolts to secure the extension blocks. Larry attached the new mounting plate for the SiTech gearbox to the extension blocks with countersunk allen-head blots. The SiTech gearbox was then secured to the mounting plate with four bolts and that also passed through the plate to the extension blocks to secure them all together. The mounting plate that would now hold the planetary gear was drilled for four 10-24 allen-head bolts to secure the new gearbox on just one end. And the L-bracket that held the interior flange bearing was modified by cutting the mounting holes on the base to the edge, making slots out of them, to give adjustment motion to the bracket.
The new design presented some interesting assembly problems as it is a little mechanically over-constrained. The extension shaft holding the drive wheel must turn freely after the two flange bearing mounts are tightened down. The outside flange bearing has two degrees of motion and the inside flange bearing has three degrees of linear motion and a degree of rotation because of the way the base of the L-mount is bolted to the base. And the result of those motions, once everything is locked down, must be the axis of the extension shaft being well-aligned (not perfectly, but pretty close) to the output shaft of the planetary gear, both encased in the shaft coupler.
Take a deep breath. The extended mounting for the planetary gear had similar over-constraints. The planetary gear is bolted to the original vertical mounting plate. We left the holes a little over-sized for adjustment, so we have two degrees of freedom there. Each of the extension blocks has two degrees of freedom and some rotation by virtue of some oversize bolt holes in the mounting plate. The new mounting plate for the SiTech gearbox has two attachment points with the countersunk allen-head bolts to the extension blocks, so the extension blocks can rotate about those bolts. And, the SiTech gearbox has four bolts which determine the location of its output shaft and which must align four bolt holes in the new mounting plate and the four drilled and tapped holes in the extension blocks. Finally, the SiTech output shaft must align very closely with the input shaft/coupling of the planetary gear.
If you skipped the last two paragraphs, there’s good news – remarkably, it is possible to find at least one solution of all of these degrees of freedom that results in shafts that rotate without binding and with all bolts in and tightened. From disassembling and reassembling it appears that there is likely a small range of such solutions, and a very large set of combinations where the last piece you are ready to tighten down doesn’t actually fit, or you end up with a misaligned part or parts. I wrote down how I did it, and ten years from now, when we have to take it apart for maintenance, I hope those notes will come in handy.
Onward to a test run in the telescope mount. The first step was to winch the rocker box off of the base because we needed to make some room for the newly enlarged azimuth drive. It was already sitting in a notch created in the plywood diaphragm that holds the center of rotation for azimuth in position relative to the support wheels. We added an additional 4 inches to the notch and winched the rocker box back in place on the base. After we settled the azimuth drive back into its place, we reinstalled the control wiring from both azimuth and altitude drives and installed the pressure adjustment bolt on the azimuth base. I felt like a mechanic working on a transmission – just needed a little more oil. Once we reminded ourselves how to work the hand pad (comes on in track mode so you can’t see motion until you change speed to slew), we took it out for a test spin.
And it spun. It moved a bit, and then, while the motor made encouraging sounds, the extension shaft with the drive wheel didn’t rotate. You may recall, this was pretty much the problem we were trying to solve with the planetary gear, but this time it seemed that our slippage was the connection between the planetary gear output shaft and the extension shaft. We benched the drive again and tinkered with it there, completely focused on the shaft coupler and how we could tighten it. No matter what we did to clean parts of any oil so the clamping action of the shaft coupler wouldn’t be lessened, or cranking on the poor allen wrench to tighten the clamp, it slipped with relatively small loads. Amid ideas flying for modifying the shaft coupler, or replacing it, we packed up and I brought the disassembled unit back to my shop.
At its maximum load, when the telescope mount is carrying the primary and secondary mirrors, the shaft coupler should have to handle 100 to 180 lb-in of torque, depending on the orientation of the OTA. My reading of the shaft coupler specs now (I’d never looked at them before, right? Shaft couplers just couple, right?) was that it should handle 165 lb-in. And we were not putting anywhere near that load on the coupler yet. I sent an email off to the manufacturer’s support to see if they had any ideas for us, and I returned to the bench to re-assemble the drive and to do more troubleshooting.
I noticed that there was new oil on the SiTech shaft and remembered that the planetary gear had a light coat on its shaft and probably had some on its input shaft coupler. So I wiped that off carefully as I reassembled it, found my second reasonably optimized assembly solution, and reproduced the slipping of the drive with a light load (my hand holding the drive wheel). I finally noticed during these trials that the visible part of the SiTech shaft on the back of the gear box was not turning when the slipping took place. Aha! The shaft coupler was just fine – it was the clutch doing its job and allowing the gear train to slip when the torque passed its threshold. Tightening the clutch adjustment plate resulted immediately in a higher load before slipping, but with the adjustment nut completely compressing the clutch spring, I could still make it slip with my hand on the drive wheel, although by applying more load.
Seven months ago, during earlier iterations, Larry had dropped off an assortment of alternative clutch springs to try out, but we had never gotten to the point of using them. Thicker wire in the spring gives a higher spring constant (more force for a given displacement in compression) and I installed the stiffest one of the bunch that was the right diameter and length. I could immediately see the difference – it was a lot more difficult to cause the drive wheel to slip. And once I tightened the clutch spring down to about half of its length, I could no longer hold the wheel by hand – it would not slip. By the way, the gap between the bottom of the drive wheel and the base of the azimuth mount is smaller than my finger diameter. My finger just barely escaped learning a hard lesson about that gap!
We now have a workable drive, although I will no longer be surprised by more iterations, and we don’t yet know if the fully loaded mount will cause slip. We’ll be testing for those limits and replacing the clutch spring with an even stiffer one if we need to. Next steps are to re-install the azimuth drive in the mount and move on to some additional testing of it and the altitude drive. We still need to decide if we want to add a planetary gear to the altitude drive, now that we know it works in the azimuth.
Postscript: Ruland, the manufacturer of the shaft coupler, got back to me quickly, and confirmed my reading of the specs on the shaft coupler we have in service (165 lb-in) and let me know that if we need to support the higher end of the torque range, that there is another unit that is a bit larger in diameter that can handle 225 lb-in, so we have somewhere to go if that element doesn’t handle the loads.