Periodic Error Correction

[leon]

Hi All

I was just wondering about this...

I have my 12" LX200 GPS aligned and it seems to function quite well, however when slewing from object to object i dosn't seem to hit the target to well, and i have to use the controller to center the object in the EP.

Now the Manual tells me that i can train the moters so that accuracy is more critical.

Well that's ok, but if i do this and use the hand controller to slew to stars and keep them centered during this procedure, am i going to knock out the alignment, or dose the scope remember all the moves, and keep the alignment settings already programed into it.

Thank You in advance

Cheers Leon

[Dennis]

Hi Leon

I cannot speak specifically on the LX200, but here is how the two GoTo systems I have used operate (Vixen Skysensor 2000 and Takahashi Temma2).

Once I have initialised the system and performed my first alignment by synchronising on a reference star, the mount “knows where it is pointing to”. If I use the motor movement keys on the hand controller, the mount continues to update its position and will still know where it is pointing to, because it computes which button(s) you pressed and for what duration and so can calculate the new position.

If however, I move the ‘scope manually by disengaging the mount’s clutches, the controller has no way of computing this manual movement, so it loses alignment and you have to start over.

Once I am aligned, if I GoTo (slew) to another star and it is not bang in the middle of the FOV, I use the hand controller to centre the star (usually in a high powered eyepiece) and then re-sync the ‘scope on that star and this usually makes subsequent GoTo’s more accurate.

Generally, my GoTo’s are accurate in the same region of the sky. If I slew across the Meridian from say, Canopus in the SE to Fomalhaut in the W, that GoTo is usually not spot on and requires me to re-centre the star and then sync. Once I have done that however, slewing around that new local region is usually accurate.

So in summary, undoing the clutches and moving the mount manually will void any alignment and you will have to start all over again. The system can only measure movements via the hand controller.

Cheers

Dennis

[Gama]

With your LX200 GPS, the software is basically the same as my RCX.
If your in Polar mode, then check your alignment, then train your drives.
If your in Alt Az, check your mount level, then train your drives.
This (Train Drives) will only work on a stationary object. So dont try it on stars. This test basically compensates for backlash and minor gear errors.
Read your manual on how to do this.
This will be your main reason to pointing error.
Others include, like i said earlier, Polar alignment, mount not flat (Use bubble level), bad calibration on your smart drv (Use factory default).

[beren]

http://www.californiastars.net/lx200gps/tips.html#

http://www.optcorp.com/ProductList.aspx?uid=383-394-398


Here you go Leon bucketloads of info to go through, you'll find the answers therein .....get those drives trained and youll be stunned with the accuracy

[rogerg (Roger)]

Leon,

If your pointing accuracy is off it's unlikely to be the fault of PEC. PEC would make a small effect but it would only be small. It's much more likely that no matter how accurate you think the polar alignment is, it's not perfect and that is leading to pointing error. Other possibilities are movement & flexing of the mount due to balance/weight.

By all means train the PEC, it will improve overall performance, but it probably won't have a large impact on the Go To accuracy.

Training the PEC does not affect your polar alignment, you don't make changes to the actual mount's mechanical position relative to the tripod during PEC, only electronic changes via the hand controller to keep the star centred.

I don't understand why Gama says not to use a star - I would suggest using a medium brightness star for PEC training. Providing it's not variable or something else strange like that.

Roger.

[Dennis]

Originally Posted by rogerg

Leon,

>snip

I don't understand why Gama says not to use a star - I would suggest using a medium brightness star for PEC training. Providing it's not variable or something else strange like that.

Roger.

We may be talking of two different tuning parameters here?

I think Gama is referring to “backlash compensation” whereas you are writing about “PEC training”.

For reducing backlash, you align on a stationary object and then press the 4 motor movement keys to see if there is too much lag before the ‘scope moves after having pressed the button. Or, the ‘scope may take off too quickly meaning that the gears could be too tightly meshed?

PEC training is where you follow the (hopefully small and smooth) wandering of a star as the mount tracks and you keep bringing the star back to the cross hairs. The mount remembers all the minor presses you make of the motor movement keys after a complete revolution of the worm and then plays this back to compensate for the wandering or Periodic Error.

Cheers

Dennis

PS - EDIT:
In my Vixen Skysensor 2000 you can set the backlash in software from a numerical value of 0 to 250 (I think?).

• A value of 0 means that there is no backlash so the software drives the motors without compensating for any potential delay.
• A value of 250 would suggest there is a lot of backlash or slack to take up, so the motors drive to take up this slack and then when the gears mesh and begin to drive the ‘scope, the system knows that real movement of the ‘scope should now be taking pace, rather than gears spinning to take up the backlash or slack.

The SS2K ships with RA and DEC values of 100 as default.

[5ash (Philip)]

Hi leon why dont you put the telescope into"High precision " on the autostar after alignment. works on my lx90 and every object is dead centre . its the only way to photograph deep sky objects.
regards philip

[leon]

Well thank you guys, i certainly have lots of info now.

Cheers leon

[Gama]

Originally Posted by rogerg

I don't understand why Gama says not to use a star - I would suggest using a medium brightness star for PEC training. Providing it's not variable or something else strange like that.

Roger.

As Dennis has pointed out, your talking about P.E.C. I was reffering to Drive Training. So in this mode the scope does not rotate (Track stars), the R.A and Dec motors are turned off. This needs to be done so a backlash compensation plus some other minor info can be calculated. So picking a star is useless as the star will drift across the eyepiece, and so pointless doing it.
You need a stationary object. I use a house about 500 meters away, and line up the window (Nice painting he has on the wall too.).


Drive Training makes a very big difference.
So it really needs to be done before you even try polar alignment or P.E.C.

[rogerg (Roger)]

I would consider both PEC and Backlash Compensation as forms of Drive Training - they both train the drives to behave differently to correct for errors.

I'm not really clear on what type of drive training you're meaning, I'm interested to know if the LX200 GPS has some additional form of drive training above and beyond PEC and backlash? It sounds like what you're correcting for in that circumstance is flexure and other variables, such as TPoint would but without having a point of reference in the sky? Is that what you mean?

Thanks,
Roger.

[beren]

Check the info contained at this link it should clarify the different training regimes for the LX200gps scopes http://www.californiastars.net/lx200gps/tips.html#

[Gama]

Originally Posted by rogerg

I would consider both PEC and Backlash Compensation as forms of Drive Training - they both train the drives to behave differently to correct for errors.

I'm not really clear on what type of drive training you're meaning, I'm interested to know if the LX200 GPS has some additional form of drive training above and beyond PEC and backlash? It sounds like what you're correcting for in that circumstance is flexure and other variables, such as TPoint would but without having a point of reference in the sky? Is that what you mean?

Thanks,
Roger.

I thought i made it clear. There is a menu in the Autostar that is specifically for... you guest it... "Train Drives", and like i said its specifically used to mesaure and compensate for errors in gears (Not just the worm and wheel) inside the drive box, etc..
The Pec is different, as you cant do a pec when the axis is rotating in the oppisite direction with gears meshing and unmeshing. As would happen when you slew to an object.
Tpoint is going overboard, but you have to train the drives at the minimum.

But, its in the manual, as i already said twice before. The LX200 does not have drive training. They started on the GPS versions on.
Download a manual and read it, or as was suggested go to the link and read up.. Theres also some other helpful info too.

[jase (Jason)]

Having read through this thread twice, I'm still not too sure where it is going...(perhaps I've miss understood this).

PEC and backlash compensation have very little to do with pointing accuracy of a telescope. PEC and backlash do however affect tracking capabilities.

To improve pointing accuracy, the telescope needs to be "modelled" i.e building a precise star map model. Such a model will take into consideration mirror flop, misalignment of polar axis, flexure etc, etc. The more points across the sky that are mapped, the better the pointing model. Automated pointing models are usually completed using software such as TPoint and Maxpoint (as RogerG mentioned). An 150-200 point model based on the GSC (Hubble Guide Star Catalogue) or Tycho II catalogue will provide an extremely accurate pointing model ensuring objects are centered on a small CCD chip everytime after slewing. Building the model can take time as the telescope slews to the 150-200 stars, but automated scripts make the task easier.

If improvement of PEC/Backlash for tracking is what you are after, you may wish to investigate software such as PEMPro.

[Dennis]

Hi Jase

Moving to a variation on Leon’s original post, I set up and tear down my 'scope each night as I don't have an observatory or permanent set up. I guess this would mean that T-Point analysis and modeling would not apply in this situation, unless you run it again each time you set up?

Cheers

Dennis

[jase (Jason)]

Hi Dennis, indeed pointing model software is generally suited to permanent installations... however having stated this, the software can also be easily utilised to accelerate precise polar alignment for non-permanent installations. In most cases eliminate the need for drift alignment.

Simply load an existing pointing model based on equipment/location set up (note: you must have already created this). Sync a few stars and the software will let you know how far off polar alignment you are. Once you've made changes in AltAz/Dec, repeat the proceedure (sync a few stars) to confirm the precise polar alignment.

You may wish to check out some of the information on such modeling operations: http://www.bisque.com/tom/Paramount/...ivemapping.asp
The link has some specific Paramount ME content, but conceptionally the details apply to other mounts.

I'd like to note, that some mounts such as ones controlled with Gemini controllers have built-in pointing model capabilities. Every time you perform an additional align on a star the point model accuracy improves. Though, it does not replace the advanced multipoint models that TPoint and Maxpoint can achieve. Also to use TPoint, ideally you should also be using TheSky as the two product intergrate tightly.

[jase (Jason)]

Sorry, to specifically answer your question. You don't need to re-do an 150-200 point model. Just load an existing one (assuming the equipment and location are same each night). You only need to sync the model.

[Dennis]

Thanks Jase, for the explanation and the link to the Software Bisque web. I just read/viewed the SB stuff - very impressive, and.... they state that Keck uses T-Point too!

Cheers

Dennis

[jase (Jason)]

Yes, most if not all of the professional observatories use a form of point modeling software. T-Point is quite common. Professional telescopes with mechanical fault tolerances kept to a minimum through advanced engineering and manufacturing processes still exhibit point issues. Software modeling simply compensates for these mechanical errors. Hence benefits can be experienced from both professional and amateur astronomers.

Most of these telescopes operate at such long focal lengths which result in a narrow field of view. This makes it difficult to ensure objects are centered on a CCD detector every time the telescope is slewed to a different object. This task would be almost impossible without some form of modeling.

[Gama]

http://www.meade.com/educational/etx...s/chap10b.html a short video on how to train drives.

The most simplest form, imagine a motor with an encoder on it. Then imagine the motor is connected to a 300 to 1 reducer gear box, then again that connected to a 180 tooth worm wheel.
This setup has a large amount of backlash. How ?, well if you can move the worm shaft back and forth, then imagine the motor would need to rotate, lets say 8 times to move the worm the same amount when its meshed. So then you are losing 8 rotations of the motor, and given it has a encoder on the motor shaft (On the Back), it assumes you have moved the scope by that amount. But in fact you havent. Now imagine this thing going on every time you slew or move or guide back and forth. This is why you train your drives. It basically will count how many pulses it will need before it should start with the real position co ordinate change.
I have obviously hyped up some figures and techniques, but i hope you now have a jist of it.
PEC will just compensate an already meshed gear system. Remember that you cannot reverse the R.A axis when your doing a PEC correction. it just slows the motor, and the earths rotation causes the star to move across your eyepiece. Plus you dont lose any pulses from the encoder, as like i said the gears are meshed in a forward motion (Sidereal).

[rogerg (Roger)]

(laughing to myself, not at someone)

This thread seems to be going all 'over the shop'

But to continue it in yet another (all be it interesting) direction:

Jase - do you have experience with T-Point and LX200's? (GPS or whatever). I have been told with someone who has experience on the subject that TPoint isn't so well suited to LX200's because once you turn the LX off, the encoder positions are not remembered and so you need to re-build your pointing model, or leave the scope on 24/7 but in land mode when not used. I believe this this was irrelevant of having parked the scope at the standard home position and was definitely assuming a permenant setup. My memory of the discussion with this person is sketchy at best, but I remember the outcome being not to spend the money on TPoint if using an LX.

Thoughts?

Roger.