VintageBigBlue.org

 

Re: PE Testing - data very "noisy"


May 11, 2009

 


----------------------------

#42807 May 11, 2009

I am going through the exercise of reducing the periodic error in my G-11 with Gemini L4 v1.02.



I believe I've reduced the PE from something on the order of 15" peak-to-peak down to 9" peak-to-peak (no PEC, yet). The remaining error is OK to guide out, but I'm hoping for a bit better.



The problem I'm having is that the PE data looks *very* noisy. As in, I don't see really any overlap from one curve to the next; it's just a random mass of squiggles that average out to ~+5/-4. I would expect to be able to overlay one worm rotation graph onto another, and have the curves somewhat line up (plus or minus seeing and other minor factors).



Has anyone else had this problem?



I am gathering the PE data with MaxIM 5.03, by keeping a log of the corrections ordered by my ST-4 autoguider. I'll post any MaxIM settings the group thinks are useful.



I am analyzing the PE data with PecPrep (just downloaded PemPro trial and get similar results).



Among adjustments I've made are:

- de-grease and re-grease the RA axis (put in white lithium grease)

- shim the worm bearing blocks (the two blocks were a bit too low, preventing the Oldham coupler from remaining as straight as possible)

- carefully adjust the worm mesh (essentially it's as tight as possible against the RA gear without causing RA stalls)

- replaced Losmandy high-precision steel worm with Losmandy high-precision brass worm

- mounted an ammeter in-line with the power supply to make sure that the mount is drawing as even an amount of power as possible (right now, it fluctuates over 0.02A or 20mA, down from ~50mA).



I will post MaxIM DL logfiles (or their EQMOD equivalents, which can be read by PemPro) and/or MaxIM/ASCOM mount/autoguider settings as appropriate.



Any ideas why the tracking data is so jumpy?



Clear skies,

Jimbo



----------------------------

#42814 May 12, 2009

Jimbo, it is pointless trying to measure the PE of your mount with guiding enabled. If you think about it, you will see that if the guiding were perfect the result would be a permanent zero! Switch the guiding off and measure the amount the reference star moves, using as high an imaging scale and as fast a sampling rate as you can conveniently manage. You should then be limited mainly by the effects of seeing, about which very little can be done, but at least you will have good data which could be lsightly smoothed if necessary to remove the effects of seeing.



Regards

John Moore

Fleet, Hants, England



----------------------------

#42815 May 12, 2009

Actually that isn't quite true, the result won't be zero, but would still need to be integrated with respect to time to see what the error was at any instant in time. The message is the same: switch the guiding off.



Regards

John Moore



----------------------------

#42816 May 12, 2009

None overlapping cycles can also be due a dominate (or at least significant)

"76 second" error phase shifting over each cycle. Does PECPrep show a peak

at 76 seconds?



Mark



2009/5/11 rushwind igda@...>

>

>

> I am going through the exercise of reducing the periodic error in my G-11

> with Gemini L4 v1.02.

>

> I believe I've reduced the PE from something on the order of 15"

> peak-to-peak down to 9" peak-to-peak (no PEC, yet). The remaining error is

> OK to guide out, but I'm hoping for a bit better.

>

> The problem I'm having is that the PE data looks *very* noisy. As in, I

> don't see really any overlap from one curve to the next; it's just a random

> mass of squiggles that average out to ~+5/-4. I would expect to be able to

> overlay one worm rotation graph onto another, and have the curves somewhat

> line up (plus or minus seeing and other minor factors).

>

> Has anyone else had this problem?

>

> I am gathering the PE data with MaxIM 5.03, by keeping a log of the

> corrections ordered by my ST-4 autoguider. I'll post any MaxIM settings the

> group thinks are useful.

>

> I am analyzing the PE data with PecPrep (just downloaded PemPro trial and

> get similar results).

>

> Among adjustments I've made are:

> - de-grease and re-grease the RA axis (put in white lithium grease)

> - shim the worm bearing blocks (the two blocks were a bit too low,

> preventing the Oldham coupler from remaining as straight as possible)

> - carefully adjust the worm mesh (essentially it's as tight as possible

> against the RA gear without causing RA stalls)

> - replaced Losmandy high-precision steel worm with Losmandy high-precision

> brass worm

> - mounted an ammeter in-line with the power supply to make sure that the

> mount is drawing as even an amount of power as possible (right now, it

> fluctuates over 0.02A or 20mA, down from ~50mA).

>

> I will post MaxIM DL logfiles (or their EQMOD equivalents, which can be

> read by PemPro) and/or MaxIM/ASCOM mount/autoguider settings as appropriate.

>

> Any ideas why the tracking data is so jumpy?

>

> Clear skies,

> Jimbo

>

>

>





[Non-text portions of this message have been removed]



----------------------------

#42819 May 12, 2009

Hi Floyd,



I would expect gearbox noise to show up as jitter on an otherwise matching curve. What I'm seeing is curves that completely don't match each other at all.



Here's a screenshot from PECprep. Note the right-hand colored window. In here, I'd expect to see several overlapping curves. What I'm seeing instead is a mess that's loosely bounded by an upper and lower bound. Is this normal?



www.jimbo.net/astro/20090510_brass_worm_AT66_ST4.jpg



I can try running in the gears.



Clear skies,

Jimbo

Newark, CA

--- In Losmandy_users@yahoogroups.com, "Floyd Blue" fblue@...> wrote:

>

> Well, part of the noise is from the gearbox as has been discussed here a lot. The other is from various things, such as the worm bearings or the worm or worm wheel to the seeing itself.







----------------------------

#42820 May 12, 2009

Hi John,



Anticipating this problem, I set it up to guide only in Dec (to capture seeing), and ran it for about 4 worm cycles, which produced this:



www.jimbo.net/astro/20090511_brass_guideonly_noRA.jpg



Zeroes would have been fine. I don't understand why I get a mess instead.



Best,

Jimbo

Newark, CA

--- In Losmandy_users@yahoogroups.com, "jcmoore_uk" j_moore@...> wrote:

>

> switch the guiding off.



----------------------------

#42821 May 12, 2009

Hi Mark,



I'm not seeing a significant contribution at 76 seconds (or any of the other "bearing-related" fundamentals). Here's a screenshot of my frequency spectrum, with just the top 25% of the errors displayed; I set the lines at ~239 (worm/1) and ~76 (the bearing thing).



www.jimbo.net/astro/20090510_brass_worm_AT66_ST4_freq.jpg



The significant spikes that I'm concerned about are the ones in the 10.2 to 34.7 range. None of those spikes align with any of the fundamentals in the range. It's like there's a whole new component of periodic error.



Any ideas?



Best,

Jimbo

--- In Losmandy_users@yahoogroups.com, Mark Crossley public@...> wrote:

>

> None overlapping cycles can also be due a dominate (or at least significant)

> "76 second" error phase shifting over each cycle. Does PECPrep show a peak

> at 76 seconds?

>

> Mark



----------------------------

#42823 May 12, 2009

Hi Floyd,



In essence, my question boils down to this.

I keep seeing other people's PE tracking data that looks like this:

f1.grp.yahoofs.com/v1/cP8JSutxtqPTboMzsMaIG5YyQd-Cza9pRSSgmZptOLE2xbjCw2RWfjRH9BSXYBHU9v2P4qV2X-ZGkWmDiCEOg3Mx-LISag/esy/captured-data-03-19-2009.jpg



But mine looks more like this:

f1.grp.yahoofs.com/v1/cP8JSu6oCQDTboMzLeGpc7B6ffESVxVYbyDln54J8OkOgJ1Lp70ZzkzRwP62t-zVobGdLlV3fcA2J5tEE7GRwKM1e7I0MQ/Brass%20worm%20and%20P5G11%20geargox%20-%20%20PEMPro/brass-worm-ejh.jpg



In the first photo, although there's an obvious Dec drift, it looks like the worm cycles all have more or less the same shape; they could be overlaid and the general shape of the PE curve would still be obvious.



In the second photo, it's a mess. None of the peaks seem to line up, although even in this photo, the overall general trend seems to be followed; near the 2 minute point, the worm is way in negative territory, near the 20-second point, it's way positive.



My logs look nothing like either of these (more like #2, but "even more so"). I seem to have a bunch of random peaks and troughs that hover near (but not quite at) 0. And, more important, the locations of troughs seem to be totally different from cycle to cycle.



Is this normal?



Best,

Jimbo

--- In Losmandy_users@yahoogroups.com, "Floyd Blue" fblue@...> wrote:

>

> Hi Jimbo,

> The gearbox has several components that it adds to the graph, three of which are 3.2s, 9.6s, 31.9s. But, each of the teeth on the nylon gear seems to add a little noise as they pass by during rotation. Plus some other stuff appears to be adding spikes too that are related to the gearbox.

>

> I can say that I had many more spikes with the Losmandy gearbox than with the MClennan gearbox. Look at this photo showing 3 graphs where there is the original worm and GB, the Ovision worm and original GB and then the Ovision worm and MC GB.

>

> tiny.cc/Lb72S

>

> The number of spikes with each mod was reduced, but you will see the original had many spikes like you are seeing.

>

> Floyd

> --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> >

> > Hi Mark,

> >

> > I'm not seeing a significant contribution at 76 seconds (or any of the other "bearing-related" fundamentals). Here's a screenshot of my frequency spectrum, with just the top 25% of the errors displayed; I set the lines at ~239 (worm/1) and ~76 (the bearing thing).

> >

> > www.jimbo.net/astro/20090510_brass_worm_AT66_ST4_freq.jpg

> >

> > The significant spikes that I'm concerned about are the ones in the 10.2 to 34.7 range. None of those spikes align with any of the fundamentals in the range. It's like there's a whole new component of periodic error.

> >

> > Any ideas?

> >

> > Best,

> > Jimbo

> >

> > --- In Losmandy_users@yahoogroups.com, Mark Crossley public@> wrote:

> > >

> > > None overlapping cycles can also be due a dominate (or at least significant)

> > > "76 second" error phase shifting over each cycle. Does PECPrep show a peak

> > > at 76 seconds?

> > >

> > > Mark

> >

>



----------------------------

#42825 May 12, 2009

I don't know why the last post did not show, but I will repeat.

I think that you indeed have some odd curves and spikes and I think it might be due to the new worm on the old wheel. Perhaps you need to run in the worm on the wheel with a drill motor at higher speed for a day to lap the worm to the wheel? This is what I would probably do.

Floyd --- In Losmandy_users@yahoogroups.com, "rushwind" igda@...> wrote:

>

> Hi Floyd,

>

> In essence, my question boils down to this.

> I keep seeing other people's PE tracking data that looks like this:

> f1.grp.yahoofs.com/v1/cP8JSutxtqPTboMzsMaIG5YyQd-Cza9pRSSgmZptOLE2xbjCw2RWfjRH9BSXYBHU9v2P4qV2X-ZGkWmDiCEOg3Mx-LISag/esy/captured-data-03-19-2009.jpg

>

> But mine looks more like this:

> f1.grp.yahoofs.com/v1/cP8JSu6oCQDTboMzLeGpc7B6ffESVxVYbyDln54J8OkOgJ1Lp70ZzkzRwP62t-zVobGdLlV3fcA2J5tEE7GRwKM1e7I0MQ/Brass%20worm%20and%20P5G11%20geargox%20-%20%20PEMPro/brass-worm-ejh.jpg

>

> In the first photo, although there's an obvious Dec drift, it looks like the worm cycles all have more or less the same shape; they could be overlaid and the general shape of the PE curve would still be obvious.

>

> In the second photo, it's a mess. None of the peaks seem to line up, although even in this photo, the overall general trend seems to be followed; near the 2 minute point, the worm is way in negative territory, near the 20-second point, it's way positive.

>

> My logs look nothing like either of these (more like #2, but "even more so"). I seem to have a bunch of random peaks and troughs that hover near (but not quite at) 0. And, more important, the locations of troughs seem to be totally different from cycle to cycle.

>

> Is this normal?

>

> Best,

> Jimbo

>

> --- In Losmandy_users@yahoogroups.com, "Floyd Blue" fblue@> wrote:

> >

> > Hi Jimbo,

> > The gearbox has several components that it adds to the graph, three of which are 3.2s, 9.6s, 31.9s. But, each of the teeth on the nylon gear seems to add a little noise as they pass by during rotation. Plus some other stuff appears to be adding spikes too that are related to the gearbox.

> >

> > I can say that I had many more spikes with the Losmandy gearbox than with the MClennan gearbox. Look at this photo showing 3 graphs where there is the original worm and GB, the Ovision worm and original GB and then the Ovision worm and MC GB.

> >

> > tiny.cc/Lb72S

> >

> > The number of spikes with each mod was reduced, but you will see the original had many spikes like you are seeing.

> >

> > Floyd

> > --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> > >

> > > Hi Mark,

> > >

> > > I'm not seeing a significant contribution at 76 seconds (or any of the other "bearing-related" fundamentals). Here's a screenshot of my frequency spectrum, with just the top 25% of the errors displayed; I set the lines at ~239 (worm/1) and ~76 (the bearing thing).

> > >

> > > www.jimbo.net/astro/20090510_brass_worm_AT66_ST4_freq.jpg

> > >

> > > The significant spikes that I'm concerned about are the ones in the 10.2 to 34.7 range. None of those spikes align with any of the fundamentals in the range. It's like there's a whole new component of periodic error.

> > >

> > > Any ideas?

> > >

> > > Best,

> > > Jimbo

> > >

> > > --- In Losmandy_users@yahoogroups.com, Mark Crossley public@> wrote:

> > > >

> > > > None overlapping cycles can also be due a dominate (or at least significant)

> > > > "76 second" error phase shifting over each cycle. Does PECPrep show a peak

> > > > at 76 seconds?

> > > >

> > > > Mark

> > >

> >

>







----------------------------

#42831 May 13, 2009

Jim,



There was a comment earlier in this thread that you need to collect data with the RA guiding corrections turned off. I don't recall seeing any acknowlegement of this and subsequent change in data. If this is true I don't think the guiding correction data can be used to glean anything really usefull about the PE and noise behavior of your dtrive train. I am not even sure if the corrections were integrated that they would give an accurate data set. Is the data you are looking at with RA corrections on or off?



Greg

--- In Losmandy_users@yahoogroups.com, "Floyd Blue" fblue@...> wrote:

>

> I don't know why the last post did not show, but I will repeat.

> I think that you indeed have some odd curves and spikes and I think it might be due to the new worm on the old wheel. Perhaps you need to run in the worm on the wheel with a drill motor at higher speed for a day to lap the worm to the wheel? This is what I would probably do.

> Floyd

> --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> >

> > Hi Floyd,

> >

> > In essence, my question boils down to this.

> > I keep seeing other people's PE tracking data that looks like this:

> > f1.grp.yahoofs.com/v1/cP8JSutxtqPTboMzsMaIG5YyQd-Cza9pRSSgmZptOLE2xbjCw2RWfjRH9BSXYBHU9v2P4qV2X-ZGkWmDiCEOg3Mx-LISag/esy/captured-data-03-19-2009.jpg

> >

> > But mine looks more like this:

> > f1.grp.yahoofs.com/v1/cP8JSu6oCQDTboMzLeGpc7B6ffESVxVYbyDln54J8OkOgJ1Lp70ZzkzRwP62t-zVobGdLlV3fcA2J5tEE7GRwKM1e7I0MQ/Brass%20worm%20and%20P5G11%20geargox%20-%20%20PEMPro/brass-worm-ejh.jpg

> >

> > In the first photo, although there's an obvious Dec drift, it looks like the worm cycles all have more or less the same shape; they could be overlaid and the general shape of the PE curve would still be obvious.

> >

> > In the second photo, it's a mess. None of the peaks seem to line up, although even in this photo, the overall general trend seems to be followed; near the 2 minute point, the worm is way in negative territory, near the 20-second point, it's way positive.

> >

> > My logs look nothing like either of these (more like #2, but "even more so"). I seem to have a bunch of random peaks and troughs that hover near (but not quite at) 0. And, more important, the locations of troughs seem to be totally different from cycle to cycle.

> >

> > Is this normal?

> >

> > Best,

> > Jimbo

> >

> > --- In Losmandy_users@yahoogroups.com, "Floyd Blue" fblue@> wrote:

> > >

> > > Hi Jimbo,

> > > The gearbox has several components that it adds to the graph, three of which are 3.2s, 9.6s, 31.9s. But, each of the teeth on the nylon gear seems to add a little noise as they pass by during rotation. Plus some other stuff appears to be adding spikes too that are related to the gearbox.

> > >

> > > I can say that I had many more spikes with the Losmandy gearbox than with the MClennan gearbox. Look at this photo showing 3 graphs where there is the original worm and GB, the Ovision worm and original GB and then the Ovision worm and MC GB.

> > >

> > > tiny.cc/Lb72S

> > >

> > > The number of spikes with each mod was reduced, but you will see the original had many spikes like you are seeing.

> > >

> > > Floyd

> > > --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> > > >

> > > > Hi Mark,

> > > >

> > > > I'm not seeing a significant contribution at 76 seconds (or any of the other "bearing-related" fundamentals). Here's a screenshot of my frequency spectrum, with just the top 25% of the errors displayed; I set the lines at ~239 (worm/1) and ~76 (the bearing thing).

> > > >

> > > > www.jimbo.net/astro/20090510_brass_worm_AT66_ST4_freq.jpg

> > > >

> > > > The significant spikes that I'm concerned about are the ones in the 10.2 to 34.7 range. None of those spikes align with any of the fundamentals in the range. It's like there's a whole new component of periodic error.

> > > >

> > > > Any ideas?

> > > >

> > > > Best,

> > > > Jimbo

> > > >

> > > > --- In Losmandy_users@yahoogroups.com, Mark Crossley public@> wrote:

> > > > >

> > > > > None overlapping cycles can also be due a dominate (or at least significant)

> > > > > "76 second" error phase shifting over each cycle. Does PECPrep show a peak

> > > > > at 76 seconds?

> > > > >

> > > > > Mark

> > > >

> > >

> >

>



----------------------------

#42836 May 13, 2009

Hello Jimbo,



Unfortunately the links you provide aren't responding at the moment, so I can't add anyhting useful to the discussion. I'll keep watching.



John



----------------------------

#42838 May 13, 2009

Hi Greg,



This set of data has RA corrections turned off:



www.jimbo.net/astro/20090511_brass_guideonly_noRA.jpg



Clear skies,

Jimbo

Newark, CA

--- In Losmandy_users@yahoogroups.com, "Greg" beneckerus@...> wrote:

>

> Jim,

>

> There was a comment earlier in this thread that you need to collect data with the RA guiding corrections turned off. I don't recall seeing any acknowlegement of this and subsequent change in data. If this is true I don't think the guiding correction data can be used to glean anything really usefull about the PE and noise behavior of your dtrive train. I am not even sure if the corrections were integrated that they would give an accurate data set. Is the data you are looking at with RA corrections on or off?

>







----------------------------

#42841 May 13, 2009

Hi Floyd,



OK, this is what I'll try next.



To be clear, you are suggesting that I drive the worm with a drill

at (16x sidereal)? or so for a day or so. Should I be doing this

with regular lubrication in place (Slick 50), or should I be using

some kind of lapping compound?



The data was noisy with the old worm, too, but the mount is fairly

new (it's a May 2007 model), so it's entirely possible that we're

still working our way through the "run in" period.



It sounds like lapping the worm and wheel together is the next step.



Thanks for your analysis.



Best,

Jimbo

--- In Losmandy_users@yahoogroups.com, "Floyd Blue" fblue@...> wrote:

>

> I don't know why the last post did not show, but I will repeat.

> I think that you indeed have some odd curves and spikes and I think it might be due to the new worm on the old wheel. Perhaps you need to run in the worm on the wheel with a drill motor at higher speed for a day to lap the worm to the wheel? This is what I would probably do.

> Floyd



----------------------------

#42847 May 13, 2009

Hi Jimbo,

No compounds just the grease. We are talking run in and not lapping with grit, but rather just wearing in the two gears so that they mesh well.



Of course the clutch would be loosened and the counter weight shaft should be installed and weighted to hold the mount from rotating while the run in is being done, so that it will not hit the tripod or pier.



The drill just keeps from over using the motor and gearbox. It will provide a RPM according to the specs of the drill. I would try to keep the RPM down to 250-300 RPM myself. A pice of tubing could probably be used as a flexible drive between the drill and the worm. Not sure exactly what to suggest there, but I would probably use tubing that would slip tightly over the shaft where the Oldham coulper fits on the worm and then use a drill bit that fit tightly in the hose and put that in the drill. This would allow you to have the drill a distance away from the mount so you could mount the drill to something to hold it steady.



So, let us say it runs at 300 rpm for instance. If you figure that the worm normally tracks at 1 rpm every 4 minutes, then 300 rpm for 1 minutes is equal to 75 minutes of tracking time. Taking that one step further, 100 minutes at 300 rpm (30,000 revs) would equal 7,500 minutes or tracking or 125 hours of normal tracking.



So I am not sure just how much time will be required, but I believe that the fellow here that ran in his OVision worm like this, used about 30,000 revs and got good results.



I would also add that it needs to rotate in the direction of normal tracking, East to West, which should be clock wise I believe. When I ran in my MC Gearbox I used clockwise rotation. But it would probably be a good thing to run it 30,000 revs in both directions, if you had a reversible drill. Then you would get both sides of the teeth worn in the same.



So, use a drill with a known RPM so you can get the revolutions of run in, figure out how many minutes or hours you wish to run it in and then time the session for each direction.



This is what I would try myself, let's see if others have different opinions on this and their reasoning for it.

Floyd --- In Losmandy_users@yahoogroups.com, "rushwind" igda@...> wrote:

>

> Hi Floyd,

>

> OK, this is what I'll try next.

>

> To be clear, you are suggesting that I drive the worm with a drill

> at (16x sidereal)? or so for a day or so. Should I be doing this

> with regular lubrication in place (Slick 50), or should I be using

> some kind of lapping compound?

>

> The data was noisy with the old worm, too, but the mount is fairly

> new (it's a May 2007 model), so it's entirely possible that we're

> still working our way through the "run in" period.

>

> It sounds like lapping the worm and wheel together is the next step.

>

> Thanks for your analysis.

>

> Best,

> Jimbo

>

> --- In Losmandy_users@yahoogroups.com, "Floyd Blue" fblue@> wrote:

> >

> > I don't know why the last post did not show, but I will repeat.

> > I think that you indeed have some odd curves and spikes and I think it might be due to the new worm on the old wheel. Perhaps you need to run in the worm on the wheel with a drill motor at higher speed for a day to lap the worm to the wheel? This is what I would probably do.

> > Floyd

>



----------------------------

#42849 May 13, 2009

Jimbo,



OK, RA corrections were off. Were DEC corrections turned off as well? DEC corrections can introduce some noise into RA measurement, as well as confuse the software that's gathering the data (or the person interpreting the data g>)



It's best to collect PE data with no autoguider corrections at all.



Regards,



-Paul

--- In Losmandy_users@yahoogroups.com, "rushwind" igda@...> wrote:

>

> Hi Greg,

>

> This set of data has RA corrections turned off:

>

> www.jimbo.net/astro/20090511_brass_guideonly_noRA.jpg

>

> Clear skies,

> Jimbo

> Newark, CA

>

> --- In Losmandy_users@yahoogroups.com, "Greg" beneckerus@> wrote:

> >

> > Jim,

> >

> > There was a comment earlier in this thread that you need to collect data with the RA guiding corrections turned off. I don't recall seeing any acknowlegement of this and subsequent change in data. If this is true I don't think the guiding correction data can be used to glean anything really usefull about the PE and noise behavior of your dtrive train. I am not even sure if the corrections were integrated that they would give an accurate data set. Is the data you are looking at with RA corrections on or off?

> >

>







----------------------------

#42850 May 13, 2009

Hi Paul,



I took this advice and last night I captured 5 cycles or so using PemPro, which as far as I know was not doing autoguider corrections in either RA or Dec.



Same problem. Data very noisy. no peaks at the same spot from cycle to cycle. No general trend line seen, just a mess ranging between +/- 5".



As an aside:

I had been doing Dec corrections before, because my polar alignment was off a little (on the order of A: 8' E: 2'). I spent some time tweaking the polar alignment before doing the no-guiding data capture.



During the capture, PemPro said Y drift was on the order of 0.002, which I can only assume means that my polar alignment has improved somewhat.



Best,

Jimbo

--- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@...> wrote:

>

> Jimbo,

>

> OK, RA corrections were off. Were DEC corrections turned off as well? DEC corrections can introduce some noise into RA measurement, as well as confuse the software that's gathering the data (or the person interpreting the data g>)

>

> It's best to collect PE data with no autoguider corrections at all.

>

> Regards,

>

> -Paul

>

> --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> >

> > Hi Greg,

> >

> > This set of data has RA corrections turned off:

> >

> > www.jimbo.net/astro/20090511_brass_guideonly_noRA.jpg

> >

> > Clear skies,

> > Jimbo

> > Newark, CA

> >

> > --- In Losmandy_users@yahoogroups.com, "Greg" beneckerus@> wrote:

> > >

> > > Jim,

> > >

> > > There was a comment earlier in this thread that you need to collect data with the RA guiding corrections turned off. I don't recall seeing any acknowlegement of this and subsequent change in data. If this is true I don't think the guiding correction data can be used to glean anything really usefull about the PE and noise behavior of your dtrive train. I am not even sure if the corrections were integrated that they would give an accurate data set. Is the data you are looking at with RA corrections on or off?

> > >

> >

>



----------------------------

#42851 May 12, 2009

Well Jim, normal is hard to find I guess. But yes, you seem to be having some irregular PE spikes and dips. But, the cause is hard to pin down. Many things could be going on here. But, I think in your case I might recommend a run in of the worm and wheel to see if the new worm can mate better with the wheel. Perhaps you could remove the gearbox and motor and use a drill to run the RA for a while. Depending on the speed of the drill, I would maybe try for 30,000 revolutions or so. I think I would leave the grease on and then clean it completely and replace the grease and try again. It just might be that the new worm and the old wheel are just not lapped together and the worm has high or low spots.

Floyd --- In Losmandy_users@yahoogroups.com, "rushwind" igda@...> wrote:

>

> Hi Floyd,

>

> In essence, my question boils down to this.

> I keep seeing other people's PE tracking data that looks like this:

> f1.grp.yahoofs.com/v1/cP8JSutxtqPTboMzsMaIG5YyQd-Cza9pRSSgmZptOLE2xbjCw2RWfjRH9BSXYBHU9v2P4qV2X-ZGkWmDiCEOg3Mx-LISag/esy/captured-data-03-19-2009.jpg

>

> But mine looks more like this:

> f1.grp.yahoofs.com/v1/cP8JSu6oCQDTboMzLeGpc7B6ffESVxVYbyDln54J8OkOgJ1Lp70ZzkzRwP62t-zVobGdLlV3fcA2J5tEE7GRwKM1e7I0MQ/Brass%20worm%20and%20P5G11%20geargox%20-%20%20PEMPro/brass-worm-ejh.jpg

>

> In the first photo, although there's an obvious Dec drift, it looks like the worm cycles all have more or less the same shape; they could be overlaid and the general shape of the PE curve would still be obvious.

>

> In the second photo, it's a mess. None of the peaks seem to line up, although even in this photo, the overall general trend seems to be followed; near the 2 minute point, the worm is way in negative territory, near the 20-second point, it's way positive.

>

> My logs look nothing like either of these (more like #2, but "even more so"). I seem to have a bunch of random peaks and troughs that hover near (but not quite at) 0. And, more important, the locations of troughs seem to be totally different from cycle to cycle.

>

> Is this normal?

>

> Best,

> Jimbo

>

> --- In Losmandy_users@yahoogroups.com, "Floyd Blue" fblue@> wrote:

> >

> > Hi Jimbo,

> > The gearbox has several components that it adds to the graph, three of which are 3.2s, 9.6s, 31.9s. But, each of the teeth on the nylon gear seems to add a little noise as they pass by during rotation. Plus some other stuff appears to be adding spikes too that are related to the gearbox.

> >

> > I can say that I had many more spikes with the Losmandy gearbox than with the MClennan gearbox. Look at this photo showing 3 graphs where there is the original worm and GB, the Ovision worm and original GB and then the Ovision worm and MC GB.

> >

> > tiny.cc/Lb72S

> >

> > The number of spikes with each mod was reduced, but you will see the original had many spikes like you are seeing.

> >

> > Floyd

> > --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> > >

> > > Hi Mark,

> > >

> > > I'm not seeing a significant contribution at 76 seconds (or any of the other "bearing-related" fundamentals). Here's a screenshot of my frequency spectrum, with just the top 25% of the errors displayed; I set the lines at ~239 (worm/1) and ~76 (the bearing thing).

> > >

> > > www.jimbo.net/astro/20090510_brass_worm_AT66_ST4_freq.jpg

> > >

> > > The significant spikes that I'm concerned about are the ones in the 10.2 to 34.7 range. None of those spikes align with any of the fundamentals in the range. It's like there's a whole new component of periodic error.

> > >

> > > Any ideas?

> > >

> > > Best,

> > > Jimbo

> > >

> > > --- In Losmandy_users@yahoogroups.com, Mark Crossley public@> wrote:

> > > >

> > > > None overlapping cycles can also be due a dominate (or at least significant)

> > > > "76 second" error phase shifting over each cycle. Does PECPrep show a peak

> > > > at 76 seconds?

> > > >

> > > > Mark

> > >

> >

>







----------------------------

#42857 May 13, 2009

Hi Jimbo,



Well, although I don't see how running in your system will help the type of

errors you are seeing, it can't hurt :) Running-in may help eliminate some of

the small random errors, but is unlikely to help the large phase shift seen in

your PE graph.



Here are the places I would look:



1. Adjust worm/gear spacing to leave a bit of room for backlash, but just a

touch.



2. Make sure the worm isn't allowed to float between the two worm blocks.



3. Check that the worm, both sides of the Oldham coupler, and the motor shaft

are all in as straight a line as possible.



4. Check that the worm rotates freely and smoothly by hand (remove the

motors/gearbox first).



5. Check the grease, a bit of dust/dirt or older, dried-up grease will cause

random errors. Check grease everywhere: on the large gear/worm, inside the RA

shaft on needle bearings, around the large roller bearing under the gear, etc.



6. Swap motors to see if the RA motor is going bad



7. Use a refractor with a locked-down focuser to test for PE: an SCT has a

movable primary mirror that often shifts under its own weight, sometimes

randomly. Make sure the camera is attached as solidly as possible to the

focuser.



These are things I would check first, before deciding on lapping or doing other

drastic things to my mount :-)



Regards,



-Paul

--- In Losmandy_users@yahoogroups.com, "rushwind" igda@...> wrote:

>

> Hi Paul,

>

> I took this advice and last night I captured 5 cycles or so using PemPro, which as far as I know was not doing autoguider corrections in either RA or Dec.

>

> Same problem. Data very noisy. no peaks at the same spot from cycle to cycle. No general trend line seen, just a mess ranging between +/- 5".

>

> As an aside:

> I had been doing Dec corrections before, because my polar alignment was off a little (on the order of A: 8' E: 2'). I spent some time tweaking the polar alignment before doing the no-guiding data capture.

>

> During the capture, PemPro said Y drift was on the order of 0.002, which I can only assume means that my polar alignment has improved somewhat.

>

> Best,

> Jimbo

>

> --- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@> wrote:

> >

> > Jimbo,

> >

> > OK, RA corrections were off. Were DEC corrections turned off as well? DEC corrections can introduce some noise into RA measurement, as well as confuse the software that's gathering the data (or the person interpreting the data g>)

> >

> > It's best to collect PE data with no autoguider corrections at all.

> >

> > Regards,

> >

> > -Paul

> >

> > --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> > >

> > > Hi Greg,

> > >

> > > This set of data has RA corrections turned off:

> > >

> > > www.jimbo.net/astro/20090511_brass_guideonly_noRA.jpg

> > >

> > > Clear skies,

> > > Jimbo

> > > Newark, CA

> > >

> > > --- In Losmandy_users@yahoogroups.com, "Greg" beneckerus@> wrote:

> > > >

> > > > Jim,

> > > >

> > > > There was a comment earlier in this thread that you need to collect data with the RA guiding corrections turned off. I don't recall seeing any acknowlegement of this and subsequent change in data. If this is true I don't think the guiding correction data can be used to glean anything really usefull about the PE and noise behavior of your dtrive train. I am not even sure if the corrections were integrated that they would give an accurate data set. Is the data you are looking at with RA corrections on or off?

> > > >

> > >

> >

>



----------------------------

#42862 May 14, 2009

Hi Paul,



I am getting better and better at taking the mount apart for re-greasing and putting it back together. Not quite ready to do it blindfolded, but you get the idea.



One of the reasons that I posted is because I've tried a number of "macro" scale changes, including:

- swapping gearboxes

- swapping main RA gear for main Dec gear (the "worm wheel")

- swapping new Losmandy brass gear for old Losmandy HP steel gear

- installed a Moonlite focuser on the AT66ED tracking scope

- swapping from AT66ED refractor to C9.25 SCT as tracking scope

- swapping power supplies from 15v@15A to 13.8v@3A and back



While my various tweaks have improved the overall PE (I started around +/- 12" and am down to +/- 5"), none of them seem to have had much effect on the underlying "noise" (which has seemed like the dominant component of the error throughout this process).



To respond to your suggestions:



6) I will try swapping the motors; one thing that worries me is that when I look at the data in PECprep, it looks like if I set the worm period to ~228s instead of ~239s (slide the slider marked "239" on the RHS center of this image: www.jimbo.net/astro/20090510_brass_worm_AT66_ST4.bmp), the curves seem to line up a little better. That makes no sense to me; I'm running at sidereal rate, I'm not seeing anything like a constant 11sec per 4min RA drift (at least, I don't think so), but this suggests that the motor is running slightly too fast?



1) I can adjust the worm for more backlash (it's adjusted for minimal backlash without getting RA stalls at this point).



2) There is no float between the worm blocks (I adjusted this out by squeezing together as I adjust).



3) Getting the Oldham coupler in a straight line was a bit of a chore. I shimmed both RA worm blocks (I used ~6 postage stamps under each) to achieve both a straight coupling (LH block) and a level worm (RH block).



4) The worm moves smoothly by hand. In addition, I put an ammeter inline with the power supply to confirm that the power usage stays fairly constant. Does anyone else do this? I have heard anecdotal evidence that leads me to believe that constant amperage (while idling at sidereal rate) is indicative of a smooth and well-adjusted geartrain. Mine fluctuates over ~20mA (between 0.17A and 0.19A) while idling (the best I could do).



5) I will probably end up doing another re-greasing run. I'll look up a re-greasing FAQ for this, to make sure I'm getting it right. Do you remove the needle bearings from inside the RA shaft? I have been through a couple of re-greasing operations, but I didn't want to do violence to get those out.



7) I have grabbed guiding data with the AT66ED (which guides at ~7"/px with the ST-4) and the C9.25 (which is at 1.18"/px with the ST-4). I did not see a reasonable difference in the amount of noise in the data either way.



At this point, I suspect that it's either:



- something wrong mechanically with the mount that is not specific to the worm, the gearbox, or the main gear (so it's the CPU, the cable, or one of the bearings, or that I have 2 bad gearboxen, 2 bad main gears, ...), or



- something wrong with the incoming power from the house (can AC noise make it through a GFI outlet, a 75dB Tripp-Lite surge protector and a ham-radio regulated linear power supply?).



...



or maybe I just don't know what I'm doing. :)



I'm trying to avoid doing anything drastic. Well, drastic and irrecoverable.



Best,

Jimbo





--- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@...> wrote:

>

> Hi Jimbo,

>

> Well, although I don't see how running in your system will help the type of

> errors you are seeing, it can't hurt :) Running-in may help eliminate some of

> the small random errors, but is unlikely to help the large phase shift seen in

> your PE graph.

>

> Here are the places I would look:

>

> 1. Adjust worm/gear spacing to leave a bit of room for backlash, but just a

> touch.

>

> 2. Make sure the worm isn't allowed to float between the two worm blocks.

>

> 3. Check that the worm, both sides of the Oldham coupler, and the motor shaft

> are all in as straight a line as possible.

>

> 4. Check that the worm rotates freely and smoothly by hand (remove the

> motors/gearbox first).

>

> 5. Check the grease, a bit of dust/dirt or older, dried-up grease will cause

> random errors. Check grease everywhere: on the large gear/worm, inside the RA

> shaft on needle bearings, around the large roller bearing under the gear, etc.

>

> 6. Swap motors to see if the RA motor is going bad

>

> 7. Use a refractor with a locked-down focuser to test for PE: an SCT has a

> movable primary mirror that often shifts under its own weight, sometimes

> randomly. Make sure the camera is attached as solidly as possible to the

> focuser.

>

> These are things I would check first, before deciding on lapping or doing other

> drastic things to my mount :-)

>

> Regards,

>

> -Paul

>

> --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> >

> > Hi Paul,

> >

> > I took this advice and last night I captured 5 cycles or so using PemPro, which as far as I know was not doing autoguider corrections in either RA or Dec.

> >

> > Same problem. Data very noisy. no peaks at the same spot from cycle to cycle. No general trend line seen, just a mess ranging between +/- 5".

> >

> > As an aside:

> > I had been doing Dec corrections before, because my polar alignment was off a little (on the order of A: 8' E: 2'). I spent some time tweaking the polar alignment before doing the no-guiding data capture.

> >

> > During the capture, PemPro said Y drift was on the order of 0.002, which I can only assume means that my polar alignment has improved somewhat.

> >

> > Best,

> > Jimbo

> >

> > --- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@> wrote:

> > >

> > > Jimbo,

> > >

> > > OK, RA corrections were off. Were DEC corrections turned off as well? DEC corrections can introduce some noise into RA measurement, as well as confuse the software that's gathering the data (or the person interpreting the data g>)

> > >

> > > It's best to collect PE data with no autoguider corrections at all.

> > >

> > > Regards,

> > >

> > > -Paul

> > >

> > > --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> > > >

> > > > Hi Greg,

> > > >

> > > > This set of data has RA corrections turned off:

> > > >

> > > > www.jimbo.net/astro/20090511_brass_guideonly_noRA.jpg

> > > >

> > > > Clear skies,

> > > > Jimbo

> > > > Newark, CA

> > > >

> > > > --- In Losmandy_users@yahoogroups.com, "Greg" beneckerus@> wrote:

> > > > >

> > > > > Jim,

> > > > >

> > > > > There was a comment earlier in this thread that you need to collect data with the RA guiding corrections turned off. I don't recall seeing any acknowlegement of this and subsequent change in data. If this is true I don't think the guiding correction data can be used to glean anything really usefull about the PE and noise behavior of your dtrive train. I am not even sure if the corrections were integrated that they would give an accurate data set. Is the data you are looking at with RA corrections on or off?

> > > > >

> > > >

> > >

> >

>







----------------------------

#42863 May 14, 2009

Hmmm. You've certainly tried a lot of things already :-(



Could it be that you just have poor seeing conditions? Seeing can certainly create a 2-3 arcsecond random errors and even long-term phase shifts in the data. Do you have an idea of how good or bad it is where you are?



I don't remove the needle bearings, mostly because they are nearly impossible to remove without specialized equipment. You can reach them fairly easily with a finger or a toothbrush. Degrease them first, then apply the grease, and pack it in by rolling the needles while pushing it in. Make sure none of the needles are stuck.



A few other ideas of what could cause a phase error between the different worm cycles:



1. Tracking a star low to the horizon where refraction can take a toll



2. Loose Oldham coupler: make sure the screws are tightened down on both sides



3. Loose motor encoder. Of course, if both motors produce the same phase shift, then that's not it.



4. Any kind of unusual friction or shifting weight, such as cables dragging.



Just going through the possibilities...



How are you gathering data from your ST4? What software do you use?



Regards,



-Paul





--- In Losmandy_users@yahoogroups.com, "rushwind" igda@...> wrote:

>

> Hi Paul,

>

> I am getting better and better at taking the mount apart for re-greasing and putting it back together. Not quite ready to do it blindfolded, but you get the idea.

>

> One of the reasons that I posted is because I've tried a number of "macro" scale changes, including:

> - swapping gearboxes

> - swapping main RA gear for main Dec gear (the "worm wheel")

> - swapping new Losmandy brass gear for old Losmandy HP steel gear

> - installed a Moonlite focuser on the AT66ED tracking scope

> - swapping from AT66ED refractor to C9.25 SCT as tracking scope

> - swapping power supplies from 15v@15A to 13.8v@3A and back

>

> While my various tweaks have improved the overall PE (I started around +/- 12" and am down to +/- 5"), none of them seem to have had much effect on the underlying "noise" (which has seemed like the dominant component of the error throughout this process).

>

> To respond to your suggestions:

>

> 6) I will try swapping the motors; one thing that worries me is that when I look at the data in PECprep, it looks like if I set the worm period to ~228s instead of ~239s (slide the slider marked "239" on the RHS center of this image: www.jimbo.net/astro/20090510_brass_worm_AT66_ST4.bmp), the curves seem to line up a little better. That makes no sense to me; I'm running at sidereal rate, I'm not seeing anything like a constant 11sec per 4min RA drift (at least, I don't think so), but this suggests that the motor is running slightly too fast?

>

> 1) I can adjust the worm for more backlash (it's adjusted for minimal backlash without getting RA stalls at this point).

>

> 2) There is no float between the worm blocks (I adjusted this out by squeezing together as I adjust).

>

> 3) Getting the Oldham coupler in a straight line was a bit of a chore. I shimmed both RA worm blocks (I used ~6 postage stamps under each) to achieve both a straight coupling (LH block) and a level worm (RH block).

>

> 4) The worm moves smoothly by hand. In addition, I put an ammeter inline with the power supply to confirm that the power usage stays fairly constant. Does anyone else do this? I have heard anecdotal evidence that leads me to believe that constant amperage (while idling at sidereal rate) is indicative of a smooth and well-adjusted geartrain. Mine fluctuates over ~20mA (between 0.17A and 0.19A) while idling (the best I could do).

>

> 5) I will probably end up doing another re-greasing run. I'll look up a re-greasing FAQ for this, to make sure I'm getting it right. Do you remove the needle bearings from inside the RA shaft? I have been through a couple of re-greasing operations, but I didn't want to do violence to get those out.

>

> 7) I have grabbed guiding data with the AT66ED (which guides at ~7"/px with the ST-4) and the C9.25 (which is at 1.18"/px with the ST-4). I did not see a reasonable difference in the amount of noise in the data either way.

>

> At this point, I suspect that it's either:

>

> - something wrong mechanically with the mount that is not specific to the worm, the gearbox, or the main gear (so it's the CPU, the cable, or one of the bearings, or that I have 2 bad gearboxen, 2 bad main gears, ...), or

>

> - something wrong with the incoming power from the house (can AC noise make it through a GFI outlet, a 75dB Tripp-Lite surge protector and a ham-radio regulated linear power supply?).

>

> ...

>

> or maybe I just don't know what I'm doing. :)

>

> I'm trying to avoid doing anything drastic. Well, drastic and irrecoverable.

>

> Best,

> Jimbo

>

>

>

> --- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@> wrote:

> >

> > Hi Jimbo,

> >

> > Well, although I don't see how running in your system will help the type of

> > errors you are seeing, it can't hurt :) Running-in may help eliminate some of

> > the small random errors, but is unlikely to help the large phase shift seen in

> > your PE graph.

> >

> > Here are the places I would look:

> >

> > 1. Adjust worm/gear spacing to leave a bit of room for backlash, but just a

> > touch.

> >

> > 2. Make sure the worm isn't allowed to float between the two worm blocks.

> >

> > 3. Check that the worm, both sides of the Oldham coupler, and the motor shaft

> > are all in as straight a line as possible.

> >

> > 4. Check that the worm rotates freely and smoothly by hand (remove the

> > motors/gearbox first).

> >

> > 5. Check the grease, a bit of dust/dirt or older, dried-up grease will cause

> > random errors. Check grease everywhere: on the large gear/worm, inside the RA

> > shaft on needle bearings, around the large roller bearing under the gear, etc.

> >

> > 6. Swap motors to see if the RA motor is going bad

> >

> > 7. Use a refractor with a locked-down focuser to test for PE: an SCT has a

> > movable primary mirror that often shifts under its own weight, sometimes

> > randomly. Make sure the camera is attached as solidly as possible to the

> > focuser.

> >

> > These are things I would check first, before deciding on lapping or doing other

> > drastic things to my mount :-)

> >

> > Regards,

> >

> > -Paul

> >

> > --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> > >

> > > Hi Paul,

> > >

> > > I took this advice and last night I captured 5 cycles or so using PemPro, which as far as I know was not doing autoguider corrections in either RA or Dec.

> > >

> > > Same problem. Data very noisy. no peaks at the same spot from cycle to cycle. No general trend line seen, just a mess ranging between +/- 5".

> > >

> > > As an aside:

> > > I had been doing Dec corrections before, because my polar alignment was off a little (on the order of A: 8' E: 2'). I spent some time tweaking the polar alignment before doing the no-guiding data capture.

> > >

> > > During the capture, PemPro said Y drift was on the order of 0.002, which I can only assume means that my polar alignment has improved somewhat.

> > >

> > > Best,

> > > Jimbo

> > >

> > > --- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@> wrote:

> > > >

> > > > Jimbo,

> > > >

> > > > OK, RA corrections were off. Were DEC corrections turned off as well? DEC corrections can introduce some noise into RA measurement, as well as confuse the software that's gathering the data (or the person interpreting the data g>)

> > > >

> > > > It's best to collect PE data with no autoguider corrections at all.

> > > >

> > > > Regards,

> > > >

> > > > -Paul

> > > >

> > > > --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> > > > >

> > > > > Hi Greg,

> > > > >

> > > > > This set of data has RA corrections turned off:

> > > > >

> > > > > www.jimbo.net/astro/20090511_brass_guideonly_noRA.jpg

> > > > >

> > > > > Clear skies,

> > > > > Jimbo

> > > > > Newark, CA

> > > > >

> > > > > --- In Losmandy_users@yahoogroups.com, "Greg" beneckerus@> wrote:

> > > > > >

> > > > > > Jim,

> > > > > >

> > > > > > There was a comment earlier in this thread that you need to collect data with the RA guiding corrections turned off. I don't recall seeing any acknowlegement of this and subsequent change in data. If this is true I don't think the guiding correction data can be used to glean anything really usefull about the PE and noise behavior of your dtrive train. I am not even sure if the corrections were integrated that they would give an accurate data set. Is the data you are looking at with RA corrections on or off?

> > > > > >

> > > > >

> > > >

> > >

> >

>







----------------------------

#42869 May 15, 2009

Let me add a #5: if you suspect that the tracking rate isn't quite what it should be, you may have a bad or failing EPROM. I would try a new one.



-Paul

--- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@...> wrote:

>

> Hmmm. You've certainly tried a lot of things already :-(

>

> Could it be that you just have poor seeing conditions? Seeing can certainly create a 2-3 arcsecond random errors and even long-term phase shifts in the data. Do you have an idea of how good or bad it is where you are?

>

> I don't remove the needle bearings, mostly because they are nearly impossible to remove without specialized equipment. You can reach them fairly easily with a finger or a toothbrush. Degrease them first, then apply the grease, and pack it in by rolling the needles while pushing it in. Make sure none of the needles are stuck.

>

> A few other ideas of what could cause a phase error between the different worm cycles:

>

> 1. Tracking a star low to the horizon where refraction can take a toll

>

> 2. Loose Oldham coupler: make sure the screws are tightened down on both sides

>

> 3. Loose motor encoder. Of course, if both motors produce the same phase shift, then that's not it.

>

> 4. Any kind of unusual friction or shifting weight, such as cables dragging.

>

> Just going through the possibilities...

>

> How are you gathering data from your ST4? What software do you use?

>

> Regards,

>

> -Paul



----------------------------

#42874 May 15, 2009

Hi Paul,



Thanks for continuing to hang in there with me. (:



It's possible that poor seeing is part of the problem, but then I would expect to see a 2-3" mess in Dec as well. The RMS of the Dec errors is usually in the sub-arcsecond range, on the order of 0.08 pixels at 7"/px, or ~0.5". I don't know whether my skies are just rock solid or whether this indicates that the errors, being random, just tend towards 0. But the Dec graph is always a *lot* quieter than the RA graph.



I shall continue to not remove the needle bearings.



Responding to your suggestions:

1) I've noticed very little difference between various parts of the sky, but for instance I've spent the last few sessions pointed at the celestial equator (within a couple of ' of it) at the meridian (within a couple of . of it), and capturing my tracking data there.



I suppose I could work at Zenith + Meridian instead, to minimize any refractive effects. I'm at 37. N latitude, so I was figuring that 53. elevation is probably good enough, plus I'm getting all the maximal polar misalignment drift...



2) I will check the Oldham coupler again. In fact, I am probably going to replace the shims. I got better at it as I was going through the process, and there's room for improvement in my neatness. Not to mention this gives me yet another opportunity to confirm that the worm has no float between the blocks and is level and engaged with the proper amount of backlash and &c.



3) I believe that I haven't yet swapped the motors; I'll try that too.



4) The only thing that I can think of re: cables is that I have all the cables run in a bundle up the mount and to the scope, strapped in place and with big loops that I've checked for strain, &c. However, among the many data cables (USB to camera, LED shutter release, ST-4 head, Robofocus cable) is one power cable (for the main camera) in the same bundle. Now I'm pretty sure that there's a ferrite bead inline in this cable. But at this point, noise induced by a power cable is not beyond my list of suspects. I just don't know how it would cause the mount to show problems. I can remove the power cable from the bundle and run it another way if anyone thinks that will help.



My tracking data collection procedure has been maturing rapidly in the past few days. (:



I run MaxIM DL 5.03, which controls both cameras and also connects to the mount (which feature I only use because it reads the Dec off the mount when it's guiding...).



Among tweaks I've made within MaxIM are:

- I've tried both "Guide Relays" and "Telescope" controlling the guiding and am currently on "Guide Relays" (the clicking is soothing)

- making sure I'm in Photo mode (not Visual)

- making sure that I'm moving at Guide Rate (not Centering)

- playing with aggressiveness (X=Y=3 normally)

- etc.



Originally, I was capturing tracking logs not only while guiding, but also while shooting images with the main camera. This causes gaps in the data (during download), and I've been counseled not to guide while gathering tracking data. OK, cool.



So I had a slight polar misalignment (on the order of 14') which made tracking without guiding at least in Dec problematic.



So I downloaded a trial of PemPro and spent an evening refining the polar alignment (about which I also have some questions, but I know it's a FAQ, so I'm researching independently) so that I could get tracking data without the guider turned on.



I got the polar alignment good enough to do this, and gathered data with PemPro (which I assume was not guiding at the time, since the ST-4 relays were not clicking). I don't really know how to read PemPro at the moment, but "Y drift" was on the order of 0.004 I think, and X was a little bigger, more like 0.1? Can't remember exactly.



Anyway, it's been cloudy for 2 nights now, so I'm waiting to see if I get some more sky and I'll grab some more data with the guide corrections turned off, and go from there.



Throughout, the RA axis still produces noisy data; +/- 5" with no matching peaks (that I can see). If I'm reading the tracking data analysis correctly, the "really big" peaks in the frequency spectrum are on the order of 1" in size. I know this isn't the whole picture, but what I'm saying is that I'm seeing 5" of noise with 1" peak periodic errors. The guider can keep up with it, but I'm weirded out that the noise just sits there, impervious to my tunings and tweakings.



sigh.



Best,

Jimbo

--- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@...> wrote:

>

> Hmmm. You've certainly tried a lot of things already :-(

>

> Could it be that you just have poor seeing conditions? Seeing can certainly create a 2-3 arcsecond random errors and even long-term phase shifts in the data. Do you have an idea of how good or bad it is where you are?

>

> I don't remove the needle bearings, mostly because they are nearly impossible to remove without specialized equipment. You can reach them fairly easily with a finger or a toothbrush. Degrease them first, then apply the grease, and pack it in by rolling the needles while pushing it in. Make sure none of the needles are stuck.

>

> A few other ideas of what could cause a phase error between the different worm cycles:

>

> 1. Tracking a star low to the horizon where refraction can take a toll

>

> 2. Loose Oldham coupler: make sure the screws are tightened down on both sides

>

> 3. Loose motor encoder. Of course, if both motors produce the same phase shift, then that's not it.

>

> 4. Any kind of unusual friction or shifting weight, such as cables dragging.

>

> Just going through the possibilities...

>

> How are you gathering data from your ST4? What software do you use?

>

> Regards,

>

> -Paul

>

>

>

> --- In Losmandy_users@yahoogroups.com, "rushwind" igda@> wrote:

> >

> > I am getting better and better at taking the mount apart for re-greasing and putting it back together. Not quite ready to do it blindfolded, but you get the idea.

> >

> > One of the reasons that I posted is because I've tried a number of "macro" scale changes, including:

> > - swapping gearboxes

> > - swapping main RA gear for main Dec gear (the "worm wheel")

> > - swapping new Losmandy brass gear for old Losmandy HP steel gear

> > - installed a Moonlite focuser on the AT66ED tracking scope

> > - swapping from AT66ED refractor to C9.25 SCT as tracking scope

> > - swapping power supplies from 15v@15A to 13.8v@3A and back







----------------------------

#42875 May 15, 2009

Hi Paul,



hmm. Certainly an option. I think I'm running L4 1.02 right now. Perhaps there's a "run more smoothly" patch in 1.04.



Best,

Jimbo

--- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@...> wrote:

>

> Let me add a #5: if you suspect that the tracking rate isn't quite what it should be, you may have a bad or failing EPROM. I would try a new one.



----------------------------

#42880 May 15, 2009

hi Jimbo im not sure if you tried swapping the gears cause this trend is long and i havnt been keeping up with it, but if you have not tried the swap you might want to try that...



robert

--- In Losmandy_users@yahoogroups.com, "rushwind" igda@...> wrote:

>

> Hi Paul,

>

> hmm. Certainly an option. I think I'm running L4 1.02 right now. Perhaps there's a "run more smoothly" patch in 1.04.

>

> Best,

> Jimbo

>

> --- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@> wrote:

> >

> > Let me add a #5: if you suspect that the tracking rate isn't quite what it should be, you may have a bad or failing EPROM. I would try a new one.

>



Contact Us
This Site's Privacy Policy
Google's privacy policies

S
e
n
i
o
r
T
u
b
e
.
o
r
g