Defeated by Orthogonality

[rogerg (Roger)]

Originally Posted by Barrykgerdes

Now set the scope up in land mode and look at a TV aerial mast some distance away, centre it and turn the telescope on and read out the azimuth. Use the RA button to slew the scope exactly 180 degrees, swing the scope over and centre the TV mast. Rotate back and forth through 180 deg. Till you get repeatable results. With a bit of practice you will be able to measure the error in arc seconds.

Barry,

For this part in your procedure, when pointing at the teTV aerial (or whatever) does the telescope not have to be pointing at 0 degrees altitude? If it can be at about 20 degrees this might be possible for me.

Thanks,
Roger.

[Barrykgerdes]

No. You are only interested in the azimuth as read from the handbox in land mode. Although it is best if you keep the altitude as low as possible and the TV aerial at least 400 metres away (makes focus easier). Use the edge of the mast at high magnification and you will be able to position/re-position to a couple of arc seconds with some practice.

A little patience will give very accurate results.

Barry

[gary]

Originally Posted by rogerg

For this part in your procedure, when pointing at the teTV aerial (or whatever) does the telescope not have to be pointing at 0 degrees altitude? If it can be at about 20 degrees this might be possible for me.

Hi Roger,

The following background may be helpful.

For the following, let us assume the mount is in an Alt/Az configuration
and is level.

Alt to Optical Axis non-perpendicularity (CA) provides a constant pointing
error residual in Az for all elevations (i.e. all values of Alt).

Az to Alt non-perpendicularity (NPAE) provides an error residual in Az that is zero
at zero Alt and its maximum value when the OTA is pointing at 90 degrees
in Alt (i.e. zero zenith distance).

In surveying on the original optical theodolites, surveyors speak in terms of
"left-face" and "right-face" readings. When you take a sighting, then rotate the
mount 180 degrees in Az and then swing the OTA back through the forks to
invert it and then re-acquire the target, you are taking a "left-face" and
"right-face" reading. Both CA and NPAE errors reverse their apparent
directions when you do this procedure.

However, to null out CA, pick a target with zero elevation. That is because
at any other elevation your error will be a function of both CA and NPAE.
Once you have nulled out CA, choose a target at a higher elevation
and attempt to null out NPAE.

If you study the last two animated graphics and the accompanying text on this web
page on our web site, it will give you further appreciation.
See http://www.wildcard-innovations.com....nt_errors.html
The material on this page forms part of some of the presentations I have given
at star parties on telescope pointing errors and analysis.

[Merlin66 (Ken)]

Barry,
As a matter of academic interest, how do you set up the fork mount level on the tripod/ bench when measuring the Dec heights?
I've been levelling the tripod with a spirit level then placing the fork mount on it.
I also adjust the OTA alignment with a spirit level across the top of the OTA when it's pointed vertically upwards. Rotate by 180 degs and re-check.

[supernova1965 (Warren)]

I was defeated just reading the name

[rogerg (Roger)]

Originally Posted by Merlin66

Barry,
As a matter of academic interest, how do you set up the fork mount level on the tripod/ bench when measuring the Dec heights?

If I understand Barry's method correctly then the telescope doesn't have to be precisely level, because you rotate it around to measure both sides, so even if the measured side is significantly lower than the opposite side, when you rotate it in azmith to measure the opposite side that side will then be at the same level the other one was. Hard to explain

Roger.

[AndrewJ]

Gday Merlin

As a matter of academic interest, how do you set up the fork mount level on the tripod/ bench when measuring the Dec heights?

I also use a similar method to Barry
The base doesnt need to be level, as long as the base of the scope
and the support holding the micrometer doesn't move.
You need to do a minimum of 4 measurements to get a correct value
but multiples are better.
Basically, I set up my extensions then point my OTA vertical
I measure the max value for the tops of the extensions
by rotating the RA 180deg
I then rotate the OTA to vertical down ( ie i move dec by 180deg )
Then repeat the process
You then average the readings for each side to get the correct delta
This eliminates any need to get the base level, or the extensions perfectly concentric to the axles.

Andrew

[Barrykgerdes]

Hi

Gary has brought up a few factors concerning the finer points of measurements which are quite correct. I did not go into them because I thought the process was complicated enough without trying to explain these factors.

One of the things I did try to stress was the difference between precision and accuracy. Many of the readings can be quite precise but the accuracy will depend on the general condition of the accuracy of the build of the telescope and the measuring equipment.

Andrewj has done a lot of work on the LX200 especially PEC correction and he is definitely the guru in this area. He also did some calculations on goto errors on telescopes polar mounted with orthogonal errors of only two or three arc minutes and how this can produce quite large errors as the observed items get closer to the poles.

The discussion on orthogonallity can be very difficult to put forward in a short answer as there are so many little things that can be missed. For most people who use these fork mounted scopes if the axes can be made within about 5 arc minutes the Gotos will be quite reasonable and almost perfect over small slews. However I have seen some telescopes with orthogonal errors close to a degree out of the box. Meade got around a lot of this by producing SMT which measures the errors at a number of points and develops a correction algorithm.

Trying to put procedures into words on how to do complicated tasks that require great expertise is almost impossible. I can tell you of many instances that cost the government millions of dollars because they wanted to save money by out sourcing work based on procedures that I tried to write as concisely and correctly as possible.

Barry

[AndrewJ]

Gday Barry

Meade got around a lot of this by producing SMT which measures the errors at a number of points and develops a correction algorithm.

Caveat here. SMT, from both code review and having a squizz at several sets of real word data doesnt work. Lots of bugs in how it saves the data.

Andrew

[Barrykgerdes]

Originally Posted by AndrewJ

Gday Barry



Caveat here. SMT, from both code review and having a squizz at several sets of real word data doesnt work. Lots of bugs in how it saves the data.

Andrew

Hi Andrew

Nice to hear from you again. I had hoped to catch up with you at the end of the month but my trip to Melbourne has been posponed.

I have never had anything to do with SMT. I don't believe in fixing errors by adding offset errors is any sort of a solution. I just mentioned it in passing.

Barry

[rogerg (Roger)]

OK, so I have made significant progress in my understanding of how to go about this orthogonality thing since starting this thread. I've now started to make something to do the job, working along the lines of Barry's method.

I bought a dial guage today and have made up a stand from metal I had in the shed to form the structure of what will hold the dial guage. See attached pictures. I'll put some bricks on the legs to hold it steady.

Could someone help me out with explanations of how to attach the dial guage to the pole in a way that I can adjust it to the correct height then lock it solid in place?

The guage has 1 inch of travel so I do wonder if I should just bolt it on hard to the pole aiming to have the pin compressed in about 1/2" and not try to have the pin "just touching" the dec axis pins I'm measuring against?

You can see the dial guage has an attachment point at the back.

I've been trying to think of how to have a leaver arm which is attached to a threaded rod or the like. Or just attach the guage directly to a vertical threaded rod some how.

Suggestions appreciated.

Thanks,
Roger.

[gary]

Originally Posted by rogerg

Could someone help me out with explanations of how to attach the dial guage to the pole in a way that I can adjust it to the correct height then lock it solid in place?

You might like to improvise yourself a couple of components. One is a clamp
that goes around the pole and can slide up and down before being fastened into the place.
The second is an articulated arm that is attached to the clamp at one end
and is fastened to the attachment point of the gauge using a machine screw
and wing nut.

For the bracket that goes up and down the pole, possibly something like
the arrangement on a TV aerial with a threaded U-bolt and a plate at the back.

For the articulated arm, a couple of pieces of narrow and thin
rectangular bracket is all you would need. Something along the lines of what
one use to find in a Mechano set. The two pieces of bracket could be
fastened together with a screw and wing nut so you can adjust the elbow to
move the dial up and down.

The first component with the bracket should provide for course adjustment in height
and the articulated arm should provide you fine adjustment.

Make sure to treat the indicator with care as it is a precision device.
Just allow the probe to touch the job gently and try and avoid the dial
being allowed to take a sudden very large deflection, both when making contact
with the job and when removing it.

[kinetic (Steve)]

Originally Posted by gary

You might like to improvise yourself a couple of components. One is a clamp
that goes around the pole and can side up and down before being fastened into the place.
The second is an articulated arm that is attached to the clamp at one end
and is fastened to the attachment point of the gauge using a machine screw
or small bolt and wing nut.

For the bracket that goes up and down the pole, possibly something like
the arrangement on a TV aerial with a threaded U-bolt and a plate at the back.

For the articulated arm, a couple of pieces of narrow and thin
rectangular bracket is all you would need. Something along the lines of what
one use to find in a Mechano set. The two pieces of bracket could be
fastened together with a screw and wing nut so you can adjust the elbow to
move the dial up and down.

The first component with the bracket should provide for course adjustment in height
and the articulated arm should give you find adjustment.

Make sure to treat the indicator with care as it is a precision device.
Just allow the probe to touch the job gently and try and avoid the dial
being allowed to take a sudden very large deflection, both when making contact
with the job and when removing it.

Roger, Gary,

Gary has just described exactly the bracket typical of a dial indicator
when used on a lathe. You can buy this type of bracket with a magnetic
base. I have one. I will try to dig out a pic of it or a link to one
on the net. Edit: This is one:
http://littlemachineshop.com/product...ProductID=1593

You can even make a right angle transfer arm for doing what you describe.
These are used for fine measurements too close to a lathe
chuck which would normally foul with the dial indicator itself.

Steve

[rogerg (Roger)]

Ahhh... thank you very much both of you. It was something as simple/obvious as using clamps that I needed.

The articulated arm - I'm sure I've seen them in bunnings or dicksmith or something somewhere ... tempting to have a look. When I asked in Toolmart about a holder today they told me one would cost about $100! .. so I didn't end up buying the guage or the holder there.

I'll have a think of how I can construct my own before going to buy one. Hmm. I have angle aluminium rods so surely can use lock nuts and the like to make something up. Concern is it not being rigid enough.

Thanks for the tips on how to handle the guage Steve.

[gary]

Originally Posted by rogerg

The articulated arm - I'm sure I've seen them in bunnings or dicksmith or something somewhere ... tempting to have a look.

Hi Roger,

I have a magnetic holder very similar to the style of the one Steve
provided the link for and the holder came with the actual dial indicator itself.

When they are on a lathe and being used to check that a job is centered,
things are a little easier because they can just be magnetically clamped
on the lathe cross slide and then the cross slide can be wound ever so
gently in until the probe just touches the job. You will have to be a bit
more careful to avoid accidentally hitting the probe too hard with your
improvised jig.

Hare and Forbes have a branch n Belmont in Perth that supplies dial holders
if you don't go the DIY route.

[kinetic (Steve)]

Originally Posted by rogerg

Thanks for the tips on how to handle the guage Steve.

Pleasure Roger,

As well as the suppliers Gary mentions, there is also Gas-Weld.
I was in there today for work to get something....
Page 7 their latest catalogue/specials...$39 bucks..magnetic.
Can't beat that.

Steve

[rogerg (Roger)]

When testing some of my equipment for the test and correction of orthogonality today I noted that, I'm quite sure, there's a noticeable difference in height from the front to the back of the fork. That is, if I wedge some metal between the RA base and the fork there is a noticeably tighter side, the side facing the control panel/south/down-hill when on wedge. If my measuring with my calipers is accurate the difference is about 1mm! The top of where the forks meet the RA base supports this, as it's not flat/same on both.

Hard to describe.

I am wondering if:
1) this is a problem, and by adjusting orthogonality with wedges of metal which will co-incidentally flatten/level this issue out, is a good thing

2) when I make adjustments I need to retain one side being higher than the other.

Another way to explain it, when mounted alt-az and telescope looking at 0 altitude, the side of the fork arms facing the eyepiece is higher off the RA base by about 1mm than the side of the fork arms on the corrector side.

Roger.

[Merlin66 (Ken)]

To hold the arms reasonably parallel I placed a length of timber ( about 25 x 150 x 500) through the handles and clamped it at top and bottom on each arm. If the bearings are machined concentric the arms should pretty well lie vertical. I ended up with the base of the fork arm parallel to the base drive unit....

AndrewJ-
With the 12" Lx200 I could get the necessary Dec rotation on the OTA (The safety stop limit the travel).................

[marki]

G'day all. watching this with great interest. Barry is it still possible to get a copy of your instructions to set up a scope? I used to have a copy but have lost it.

Mark

[rogerg (Roger)]

Well, I have much to report...

On Sunday, after a week of preparation, I attempted to measure and adjust the orthogonality of my LX200. I had limited success but progress none the less. I have attached a bunch of pictures to show some of what went on.

It took me the whole morning and until 3pm in the afternoon to un-wire and dismantle the LX200 in the observatory to a state that I would be able to take it off the Pier, and collect all necesarry tools and pieces of equipment together on the patio where I would do the alignment along with my measuring 'jig' for the fork height, etc. Then at 4pm with the help of a friend I took the LX off the pier and put it on the tripod under the patio. Then from 4 until 7 we worked on measuring and adjusting the alignment.

I'm quite confident the measurements made of the fork height and the OTA alignment were accurate. The readings I was getting were very repeatable and stable thoughout the many measurements. I used a spreadsheet to record readings for the fork heights - Left up & down, Right up & down. The measurements were averaged and compared.

I followed this procedure:
1) measure fork height
2) measure OTA alignment
3) adjust OTA alignment
4) adjust fork height
5) measure fork height
6) measure OTA alignment

After step 1, measuring the forks I had 0.58mm difference in fork height. Not too bad I thought, surprisingly close really. I was expecting much more. I moved on to measure the OTA.

Measurement of the OTA was showing 24 arc minutes of error. Reasonably large amount I thought, but I don't have anything to guage it by to know how big or small that is.

Then I went to make an adjustment and here is where I ran in to problems. First up, it seems both my OTA and my forks are adjusted to the extremes of the directions they would need to be further adjusted in. The way it's all adjusted it looks to me like if it was all losened off and reset it could be brought in to better alignment but because one's at an extreme the other is. Just a gut feel, from looking at it, it's like it's all twisted but if straightened and started from scratch would be fine.

Anyway, I tried to make an adjustment to see if I could go any further or straighten anything.

I first tried adjusting both OTA and fork (one at a time) in the direction desired. I losened off the required screws and tried moving it - nothing, no movement at all. not an iota of movement (that I could tell). Odd I thought. In the case of the forks I could see the gap now present between fork arms and base, so surely I should have been able to move them? I tried in the other direction, in the direction that there was plenty of movement in the screw slots - still no budge. No matter what I tried, with quite heavy force, I couldn't move either the forks or the OTA adjustments in any direction, even when screws were losened. I stll don't know why.

However ... after trying all that and tightenning it all up again, I re-measured, and interestingly there was a difference. Not sure exactly when the movement occured, but I now had....

Fork error of 0.24mm in the opposite direction to what the 0.58 was previously, and an OTA error of 28 arc minutes.

So, if my measuring is correct, I now have less fork error but slightly more OTA error. In reality I think I changed very little and most of the difference measured between before and after is probably in error of measuring, although I am quite confident my measuring was quite accurate.

I wouldn't be surprised if the differece beteween before and after is purely due to me losening and re-tightening everything and in no way a consequence of my attempts to make adjustments when losened. And I wouldn't besurprised if it's changed slightly since lifting the beast back on to the wedge.

For now the LX200 is back up on the pier but not yet all fully set up again (permenant cables, cameras, focuser, etc).

Back on the pier some slews and visual work last night showed the goto accuracy was OK but not brilliant. I was glad to see I hadn't thrown it completely off, it is still in the relm of "somewhat like before". I haven't done any exhaustive testing, just moving around a few objects like M42, NGC 1365, M79, Eta, etc. And I haven't tweaked the polar alignment since putting it back on the wedge so that will be off somewhat after taking the scope off and putting it back on.

I'm letting it sit there for a few days while I consider if I will try anything more regarding orthogonality before I go to the trouble of re-attaching everything. I think chances are that I'll just re-attach everything and see how it goes, perhaps reserting to TPoint, which I have but prefer not to use.

Input, comments and constructive criticism of my technique greatly appreciated.

Roger.

PS.... so perhaps after all that I'm still defeated by orthogonality