Last updated: 8 June 2008
Sent: Thursday, June 5, 2008 10:54:55 From: Chris Erikson (Chris.Erikson@insitu.com) First off I would like to think you for the invaluable contribution your site makes to ETX owners. You have created the gold standard gold mine of ETX/DS info. I've been visiting for months and still haven't read everything, and have learned a heck of a lot. Secondly I have a contribution to make below the hash marks you'll find the contents of a thread from CN concerning my trials and tribulations with GOTO runaway and uncommanded speed changes during drive training. Info from your site gave me the confidence to open it up, then I ran into the fact that many of the articles concern the last revision of the ETX base, which has smaller gears and an entirely different worm setup and motor placement than the newer ones. You do have one link to some pictures of the guts of an late model ETX70 with the plastic gears, here is info on what I ran into on my ETX60BB with the large gears, my problems and frustrations, then how I resolved the issue..part of which I had inadvertently caused. It's wordy but informative I hope. My desire is to contribute to maybe helping some other poor sod figure out their troubles. ------------------------------------------------------------------------ Part 1) The problem Hi folks The 497 handset I picked up was working well but now the system has developed a couple problems. The other night my GOTO's started failing big time, it would get close then during what is usually the 'refinement' phase where it zeros in, suddenly the scope just takes off at high speed and slews all over the place. I pulled the scope apart (1st time dissassembly) to check the encoders and found grease everywhere in there, they really slop it on. I didn't see any in the encoder wheels but I carefully used a fine q tip to clean alt and az encoders anyway. Spun them each again to check that no cotton strands were left on them, they look clean. Yet now I have two problems. I reassemble and problem number one during reassembly is I am not sure how tight to preload the big plastic nut which retains the RA axis and is underneath the roller bearing and outer locknut that the RA bolt screws into. I suspect I have it too tight as now the base seems 'sticky' in RA when locked, but if it's any looser I can't find the spot where the RA knob will lock at one position and allow free smooth motion in RA on the other. If I set it for smooth free motion, it will not lock. If I preload enough to get a lock, it's not smooth when I unlock it. I finally find a spot where it's kinda smooth when the RA is not tightened, and locks when it is. Now I go to 'calibrate motor' and it fails. Try again, it works. Try again, it fails. Try again, it works. I give up when it passes and proceed to train drives. I start in RA training, center up the target, it asks for > to place target back in center...and halfway through the slew to recenter the target..the speed changes and I get runaway. What is going on and what is the technique for getting this stuff right? Part 2) The resolution I believe I solved this problem. It appears that the issue was a combination of two factors. First factor...the inner large plastic nut which is the primary retainer for the aluminum shaft RA axis was too tight. This is the nut with the castellated (sp?) top (slotted top as used with cotter pins in other applications). The RA bolt threads into large threaded washer type nut which traps and retains the slots in the lower plastic nut. Anyhow, I think this was too tight. Without the RA/AZ bolt in place, I tightened the plastic nut until the axis wouldn't move, then slowly backed it off in 1/8 turn increments so that it held the base on firmly, but didn't stick when rotating and turned smoothly. NOTE: The plastic nut comes with locktite applied. I have not reapplied it and if it stays adjusted, I'm going to leave it off. Second factor. When I had it apart the first time to check the RA/AZ encoder wheel, it seemed to me the metal spring clip which serves as a pressure spring for the worm gear was too loose. I carefully tweaked it wider (more pressure) a mm or so. I think this was a big mistake. I returned it to about where it was when I got in there the first time. Apparently caution is the word when you think you are smarter than the factory, in some cases at least. I relearn the old saying the hard way...if it ain't broke, don't fix it. So after returning the tension 'spring' to it's prior place, and spending some time experimenting with the plastic nut torque, I was VERY pleased to get no motor errors on my first Autostar boot while on the bench after pulling my hair out for a week on this thing. I think the problem was twofold..motor calibration and motion sensing probably relies on consistent "ambient" resistive torque due to friction in the non ball bearing system here, so it's very sensitive to preloading. In addition, having the preload on the worm gear too high probably messes up the same kind of thing throughout the gear train. I think the system was sticking and then breaking loose throughout the bearings/gear system and this was confusing the processor. A DC servo motor system relies on rate measurements for position estimation. This means spikes in rate of velocity change due to high stiction will totally hose up the math used to figure out position, since they must use derivatives of velocity in order to calculate position. I think my speed changes were due to the processor thinking it was driving constant speed or whatever, but since it's math was based on sensing speed which was being skewed by drive train issues, the real speed was not what it thought. One way I was able to tell immediately if these combined adjustments were doing anything is that once out of time/date mode, your scope should immediately settle to it's normal beeps and chirps as it does default position tracking, even if you are not yet aligned. When it was screwed up, I would get runaway or go straight into motor cal mode immediately after time/date settings. When I had it right, that did not occur and I knew something had changed for the better. Lesson here is...if you're ever getting runaway at any speed or go straight to motor cal after time/date, something is getting sideways in your drive train or bearing setup. Part 3) A couple more things While I had it on the bench and once I got it tuned right so it would boot normally without runaway or instant motor cal mode, I did a drive training on the bench. Normally you are advised to train drives in the mount style you will be using, and this makes sense. I use polar mode with my PVC pier (see ghetto ETX pier thread). However, since I wanted to see if I could get through the entire process before I left the shop, I did drive training on the bench in ALT/AZ position anyway, indoors. Since my shop is not big enough to reach close focus on something inside, I couldn't look through the scope for the training process. I decided to find a good point target to use in conjunction with the laser pointer. What worked out *great* was a corner of the room near the ceiling. It turns out that where two walls and the ceiling come together makes a pretty accurate target for the green dot. When you're off center, the laser is smeared and not round, and the small off axis glow around it isn't round either. When you're aimed smack dead center at the corner, the dot is a perfect circle centered in the perfect circle of the off axis glow. Even the tiniest bit of off center, and it's obviously not right. So I ran through the process without a hitch (yee ha!), then dragged the whole works out to the pier for the do or die real world test. Usually I use a barlowed 9mm and a distant target for drive training, in the correct mount orientation. here I used the laser and a corner of the shop at most 8m away, in the wrong scope orientation. While setting up and getting located at polar home, I was ready to be dissapointed. After diddling with this problem for over a week I was getting really cranky and in spite of the obvious behavior changes on the bench, I was prepared to restrain myself from throwing the whole works in the dirt when it failed again. When combined with my wrong orientation drive training indoors, the best I was hoping for at least no runaway/speed changes and the need for a drive retrain in polar mode. Well, the results were more than satisfying. Date, time, two star, check. Polar home, check. Laser On. GOTO Arcturus. Whine whirr chug, slew, slow slew, beep. A couple degrees off Arcturus, no motor fault, no runaway! Not bad. Hopes rise. Center Arcturus, enter. Next, Vega. Whine whirr chug, slew, slow slew, beep. Laser looks danged close. Peek through the 9mm...there it is in the field! Woohoo! Center it up. Enter. Alignment Successful!(Pole error 27' ^, 47' <. Ok, whatever. I'll deal with that later. I love this report function, though. 497 rocks) M13...yup. inner 50% of field. Ring Neb...bingo. Inner 50%. User ecstatic. I play around and determine I'd be a field off after a GOTO transit of the whole sky, but GOTO'd to objects within a quandrant or so it's always in the field. Leave it for an hour and object remains pretty danged close to where you left it. For now it actually works a little better than when it came out of the box. Suuuweeeet! I was pretty late by now and I didn't feel like dragging out the DSI and associated junk, so I trotted out the Z12 and spent a pleasant hour using the laser pointer and ETX GOTO in parallel with the Z12 as a finder system for it. M63, M94, M5, a little of this a little of that. M56 is sorta underwhelming compared to the other great globs for this season. M63 and M94 are pretty danged cool in the Z. NGC6543 is danged bright! Wow in the Z12 and easily visible in the ETX, though dinky. All in all the most satisfying night out in a while even though viewing through sucker holes. Actually getting something like this fixed and working well is a pretty good feeling. The indoors drive training using the GLP appears to have worked pretty nicely and I will not hesitate to use it again when necessary. That's it. Hope it's useful. Thanks for the great site. Best Regards, Chris EriksonMike here: Just to confirm that before you did the repair, did you try a RESET, followed by a CALIBRATE MOTOR and TRAIN DRIVES on the AutoStar? Sometimes doing the RESET clears a runaway problem.
Yup, a general reset was the first thing I tried. Reset, calibrate, train. Then I went back to my 494 handset, and it had the same behavior. I'm still not quite certain how I didn't get the plastic nut torque right the first time as I was pretty careful, but I think my extra caution on it the last time around combined with fixing the worm tension did the trick. Chris Erikson
Go back to the Tech Tips page.
Go back to the ETX Home Page.