TPAS sampling strategy answer for Mike

Mike Van Lysel wrote:
Hi Gary – Thanks for the reply.

I certainly understand the importance of the azimuthal reference points provided by the initial 2-star alignment. And I suspected that, as I believe you say, the IE TPAS-term benefits from different azimuthal-samples. However, it wasn’t clear to me that there is any other azimuthal component to the modeling.

To provide a little background; I am working on a telescope (a DOB with the AZ-encoder mounted to the azimuth pivot bolt) that has ALT-issues. For one thing, the ALT-bearings are eccentric. I, just like everybody else who turns to TPAS, am looking for term(s) that are stable over time. I’ll determine these carefully and save them for future use and then resync the model with a few star samples before each observing session. So far, it looks like ECEC is significant and stable for sure; less certain but promising are NPAE and CA.

But I encountered something unexpected in my TPAS runs (I have performed 6 on this scope). The scope has a fundamental raw rms residual of about 20 arcminutes. I started out doing all-sky runs but these produced modest results. I tried various combinations of terms but the best case improvement only brought it down to about 10 minutes. Then one night I did a restricted sampling in one quadrant of the sky and the TPAS model brought the raw 20 minute residual down to 2.5 minutes. I confirmed this behavior the next observing session; a sample that was pretty much pure ALT brought 20 minutes down to 2.5 minutes. Something about the all-sky samples produces worse results than the more limited samples (I have some ideas about what, but maybe that’s for another time).

So that’s the genesis of my posted question. In the perfect world I would just go out and do a lot more TPAS-runs, but in the real world of Wisconsin weather I am trying to maximize the productivity of the chances I get by asking for help here. For my next attempt I was thinking of sampling 4-stars of equal ALT spaced at 90-degrees in AZ, then adding an ALT-ladder of about 15-stars at similar AZ equally spaced in ALT between 80 to 20 degrees (in the northern hemisphere this time of year there’s a very nice ALT-ladder to the NE in the early evening). Is this strategy smart or dumb?

Thanks much,

Mike Van Lysel

Hi Mike,

Thanks for the post.

As mentioned in my previous response, there are additional terms that are always modelled but are hidden from you.
These terms include azimuthal components. However, because the user does not establish any consistent zero reference point in
azimuth, these hidden terms aren't saved in NVRAM. A short sampling run re-establishes them on a subsequent observing session.

With your eccentric bearing, you may well benefit with either or both of ECEC and ECES.

Typically you might find either one of NPAE or CA to be of benefit as well.

It sounds like from your sampling run that you might have something going on in azimuth as well.
Chances are you performed the Daytime Encoder Test but if not I highly recommend you do so in the azimuth axis.
This might then betray something slipping. For example, if your azimuth pivot bolt is secured to the base of the ground
board like on the Obsessions, try and have a look under there and make sure the plate is still rigidly affixed.
Also make sure that the azimuth tangent arm is ideally floating a little. Look out for if it is becoming foaled on the base
of the rocker.

Nothing will trow out a pointing model faster than a misidentified star. Be sure to periodically review the sampled stars
in the REVIEW DATA sub-menu. You can change it to display between raw and fitted residuals. Have a look a the raws
and check for any obvious outliers. Don't delete a correctly sampled star simply because it has a larger error residual.

I recommend you sample the whole sky including with changes with azimuth. Otherwise you are biasing your sample
to include only stars that give you a small RMS.

Hi Gary: Yep, I’ve performed a Daytime Encoder Test. This scope started out with awful AZ encoder reproducibility, but with work I now have 360-degrees of revolution registering as 360-degrees on my AN and all of the results I discuss were obtained under this condition.

The important observation out of my data is that the scope has a raw-rms of 20’ whether I do an "all-sky" sampling run or a “restricted-sky” sampling run. IOW, it is NOT the case that I get a small model-rms with the restricted-sky run because I started out with a small raw-rms. Rather, BOTH all-sky and restricted-sky samples start out at 20’ raw-rms. The best-case model-rms values are 10' for the all-sky sample, 2.5' for the restricted-sky sample. That’s why I suspect going forward I’ll get a better determination of model term values with the restricted-sky sampling.

Regards,
Mike

Hi Mike,

Notwithstanding that your altitude bearings are eccentric, have you done the red laser test to ensure that the altitude encoder is centred, as best as can be centred, on the altitude axis of rotation? I found this helped pointing accuracy enormously in the pre TPAS days. It also helps post TPAS as there are less errors to be corrected by TPAS.

If you haven't done it I can walk you through it.

Attached is Scott Tannehill's PowerPoint presentation on the method he used for locating the alt axis on a Dobsonian.

Filename Version 4.0 Locating ALT axis on Dobsonian.ppt

Description Locating Alt Axis on a Dobsonian by Scott Tannehill

Filesize 302 Kbytes

Downloaded 177543 time(s)

[Disk] Download

ausastronomer wrote:Hi Mike,

Notwithstanding that your altitude bearings are eccentric, have you done the red laser test to ensure that the altitude encoder is centred, as best as can be centred, on the altitude axis of rotation? I found this helped pointing accuracy enormously in the pre TPAS days. It also helps post TPAS as there are less errors to be corrected by TPAS.

If you haven't done it I can walk you through it.

Hi John! Rosalie says Hi! too. Hope you are well. We are traveling to NZ this spring (I'm taking the travel scope that's the subject of these posts). We haven't yet visited all the places you recommended for our last trip!

Thanks for the kind offer to help. I don't know the red laser test, but it looks like Gary posted a how-to guide so I'll take a look at that.

I discovered this problem because I could see that the ALT-encoder was moving as I moved the scope in altitude. I set out with paper and protracter to find the "right spot" but discovered there wasn't one right spot. But, as you say, there is probably a "best" spot.

Many thanks,
Mike

Gary - Thanks much for the PowerPoint! I'll take a look.

Mike

MikeV wrote:

ausastronomer wrote:Hi Mike,

Notwithstanding that your altitude bearings are eccentric, have you done the red laser test to ensure that the altitude encoder is centred, as best as can be centred, on the altitude axis of rotation? I found this helped pointing accuracy enormously in the pre TPAS days. It also helps post TPAS as there are less errors to be corrected by TPAS.

If you haven't done it I can walk you through it.

Hi John! Rosalie says Hi! too. Hope you are well. We are traveling to NZ this spring (I'm taking the travel scope that's the subject of these posts). We haven't yet visited all the places you recommended for our last trip!

Thanks for the kind offer to help. I don't know the red laser test, but it looks like Gary posted a how-to guide so I'll take a look at that.

I discovered this problem because I could see that the ALT-encoder was moving as I moved the scope in altitude. I set out with paper and protracter to find the "right spot" but discovered there wasn't one right spot. But, as you say, there is probably a "best" spot.

Many thanks,
Mike

Hi Mike,

Say hi to Rosalie !!

Sue and I will be in New Zealand ourselves for the first 2 weeks of December. We sail on Celebrity Solstice on 20/11 for 12 Nights out of Sydney and depart the ship in Auckland on 2/12. Sue's parents will drive up from Napier to Auckland in their Motor Home and meet us off the ship in Auckland. We will then spend 2 weeks travelling in the North Island with them, before flying home about 14/12. We won't do too much travelling around as her parents are getting on, we don't have that much time on this trip and the 4 of us have all done countless trips around NZ previously. We will revisit some favourites like Taupo and Rotorua and maybe head up to the Bay of Islands for a few days. I never get tired of revisiting places like Wai-O-Tapu https://www.waiotapu.co.nz/ and Huka Falls http://www.hukafalls.com/ which are within about 50km of each other on the highway between Rotorua and Taupo

Firstly, if you can't adjust the altitude encoder position in both axes you don't have any options that don't involve a lot of work on the mount. On a scope like an Obsession (which is what is pictured in Scott's powerpoint) it is very easy. Scott's powerpoint is very detailed and actually makes the process sound a lot more complicated than what it is.

Let me explain it simplistically:-

I prefer to do this at night in a position where I have an exterior light, with the switch in a handy position. You turn the light off when moving the scope through its altitude range, which makes it easy to see the movement of the laser dot and then turn it on to make the adjustments under good light. I use my normal red dot laser collimator. I set the laser up about 2 or 3 metres from the scope, which makes it easy to move around the scope without kicking the laser rig over and having to start again. You want to set the laser up so that the laser axis is "very roughly" in line with the altitude encoder axis. It doesn't need to be that accurate, as you are working off the point, not the plane. You also however, don't want the laser firing from a 45 degree angle, as the red dot will elongate to an ellipse. I usually set the laser on a milk crate or chair or something around that height and then a small cushion with a heavy book on the top. I then "wedge" the laser between the book and the cushion. The cushion allows you to "wiggle" the laser and make slight adjustments to where it points to get it hitting the dead centre of the encoder axis shaft, but able to retain its position when it is set.

I then loosen the adjustment screws and nut slighty on the encoder bracket so that the encoder remains in place without moving, but you can move the encoder yourself, to make the adjustments as required. This saves tightening and undoing the screws and nut 400 times, while you play with it. Move the scope through it's full altitude range and you will see the laser dot move as the scope moves through its full range. A little trial and error will soon show you where you need to move the encoder to make appropriate corrections. Keep going with trial and error until the laser dot remains fixed on the centre of the encoder shaft as the scope is moved through its full altitude range. It usually takes me about 10 to 15 minutes only from start to finish to get the altitude encoder spot on. I do this as a matter of course every time I remove the altitude encoder on any of my scopes. It helps pointing accuracy a lot, particularly for those that don't do a TPAS run before an observing session.

Cheers

Attached is a one page PDF document that had been uploaded by user "obsessionowner" on the Argo Navis Yahoo Group entitled
"How to Center the Encoder Bracket on the mirror box" by Charlie Starks and Jean-Paul Richard.

Filename How to Center the Encoder Bracket on the mirror box.pdf

Description How to center the encoder bracket on the mirror box

Filesize 6 Kbytes

Downloaded 119531 time(s)

[Disk] Download

ausastronomer wrote:
Firstly, if you can't adjust the altitude encoder position in both axes you don't have any options that don't involve a lot of work on the mount. On a scope like an Obsession (which is what is pictured in Scott's powerpoint) it is very easy. Scott's powerpoint is very detailed and actually makes the process sound a lot more complicated than what it is.

Hi John - Thanks for the amplification on the red-laser method.

Unfortunately, I can not easily move the ALT-encoder on this scope. But the first task is to determine if there's a substantially better location and therefore worth the effort so I appreciate your help.

Cheers,
Mike

wildcard wrote:Attached is a one page PDF document that had been uploaded by user "obsessionowner" on the Argo Navis Yahoo Group entitled
"How to Center the Encoder Bracket on the mirror box" by Charlie Starks and Jean-Paul Richard.

Thanks again, Gary!

Filename	Version 4.0 Locating ALT axis on Dobsonian.ppt
Description	Locating Alt Axis on a Dobsonian by Scott Tannehill
Filesize	302 Kbytes
Downloaded	177543 time(s)
Download

Filename	How to Center the Encoder Bracket on the mirror box.pdf
Description	How to center the encoder bracket on the mirror box
Filesize	6 Kbytes
Downloaded	119531 time(s)
Download