64 replies to this topic

[johngwheeler]

I have a Takahahsi EM-200 Temma 2 mount and have been using SkySafari or Cartes du Ciel for GoTo control. However, both of these applications only support a simple one-star alignment, which I have found to be *very* sensitive to accurate polar alignment. Even after doing a plate solved PA, I find that large slews across the sky fail to go-to accurately at medium to high power - accuracy is probably only about 1º which might be just about OK at low power in a C8, but not at >150x. 

 

So I'm looking for options for multi-star goto alignment (at least 2 stars!), which I hope would improve goto accuracy.

 

I'm aware of an iPad application called Astromist, but this looks a little bit old-fashioned and I'm concerned that it may not be supported long-term. I mostly use a laptop, so a PC or Mac solution might be preferable to an iPad app. However, it isn't very expensive, so might be a reasonable option even if I only use it for alignment.

 

Does anyone know of any other planetarium software that support a multi-point goto alignment model with the EM-200? I have a feeling that the The Sky X can do this, but I'm not sure whether it requires the (relatively expensive) T-Point add-on.

 

Any suggestions?

 

To EM-200 owners, do you also find that the go-to accuracy with one-star alignment is not that great compared to other mounts that use multi-star alignment?

 

 

Thanks,

 

John

 

 

 

 

[Ken Sturrock]

Hey, John, as we've discussed - I do find that the GoTo on the Temma is not brilliant but does generally work OK. As you say, it is pretty touchy. I don't have the polar alignment challenges that you have in the southern hemisphere and do also use plate solves to align the GoTo system. I also use a "plate solve enabled" slew function (Closed Loop Slew in SkyX) to improve accuracy for long distance slews.

 

SkyX with T-Point would help improve GoTo accuracy on the Temma, but SkyX alone would also give you pretty darned accurate slews with just the Closed Loop Slew function. T-Point would require that you either rebuild a model before you start imaging, or make sure that your setup doesn't meaningfully change and take a smaller number of images (maybe even one?) to update the model. If you're doing traditional DSO imaging on one or two targets a night, it's probably not worth the hassle of T-Point for a Temma, especially on a mobile setup.

 

My only other suggestion is to align the GoTo on a star very close to your target. I find that my Temma is pretty accurate when slewing short distances, but grows less accurate the farther it has to slew - especially if there's a meridian flip since the mount doesn't grok cone error. I hate to say it because it's kind of a heresy on a Takahashi, but have you tried some sort of assisted polar alignment routine? I'm sure the Polemaster zealots will show up here any second, but there are also routines built into PhD or even just taking a stab at drift aligning the thing.

 

Sorry it's still giving you grief.

[orlyandico]

I am in Singapore where there's no view of Polaris either.

 

My "solution" is simple - when doing a long slew e.g. with Sky Safari, instead of slewing to the object, I slew to a bright star near the object.

 

As expected, the slew will be off.  Then I center the star with the finder (and then the main scope) and resync Sky Safari.

 

Now, the short slew to the nearby object of interest will be pretty accurate.

 

It's annoying, but works.  Don't hold your breath waiting for a multi-star solution. You could use Cartes du Ciel and Chuck's ASCOM driver. That will allow two-star aligns. But it needs a PC.

[syscore]

If I were doing it, I would use PemPro's drift align, but seeing that the Tak has no PEC, you would get only half the value out of PemPro.

 

Ken, assuming a PAE under 5', how close are the GOTO's?

 

John, do you have SGP? It does sound like closed loop is the way to go.

[syscore]

I am in Singapore where there's no view of Polaris either.

 

My "solution" is simple - when doing a long slew e.g. with Sky Safari, instead of slewing to the object, I slew to a bright star near the object.

 

As expected, the slew will be off.  Then I center the star with the finder (and then the main scope) and resync Sky Safari.

 

Now, the short slew to the nearby object of interest will be pretty accurate.

 

It's annoying, but works.  Don't hold your breath waiting for a multi-star solution. You could use Cartes du Ciel and Chuck's ASCOM driver. That will allow two-star aligns. But it needs a PC.

That is actually a good option, and a common option not too long ago. It is still an option on Celestron hand controllers, called "high precision slew", but because they have a cone error measurement, slews are not that far off across the sky, so it isn't a much used option anymore. However, when I use my 11" with my ASI178, which is essentially the same as using a 6000 mm focal length, I have to often use the high precision slew option.

[orlyandico]

it's pretty hard to get sub 5' PA without Polaris visible. You northerners don't know what a good thing you have.

 

i can only get that level of PA if I drift align with PEMPro, which takes forever (well, 20-30 minutes). allegedly the polar scope should work in the southern hemisphere, although I've never seen or tried it.

 

There's an even more idiotic/simple method that I use.

 

Because my EM-11 is a T2Jr (which slews very slowly..) if I have to do a really long slew that requires a meridian flip (I know the slew will be inaccurate anyway and it takes forever) I just declutch and manually move the mount to the other side of the meridian.  Then I center it on a bright star near my object of interest, then resync Sky Safari.

 

Voila.

[syscore]

I was wondering if I was using one here, where I can see Polaris. In my reviews (of forums), I didn't see goto's as being that bad, and since an EM-200 is used with refactors mostly, you have a pretty wide FOV to miss. You have a good point regarding slow slews, but this would be an imaging mount, and I would point (like you do) and then sync, near my object, and that would be it for the night.

Edited by syscore, 20 October 2016 - 10:18 PM.

[syscore]

Or is the main problem, needing to sync again when you slew to the opposite meridian? 

[johngwheeler]

I am in Singapore where there's no view of Polaris either.

 

My "solution" is simple - when doing a long slew e.g. with Sky Safari, instead of slewing to the object, I slew to a bright star near the object.

 

As expected, the slew will be off.  Then I center the star with the finder (and then the main scope) and resync Sky Safari.

 

Now, the short slew to the nearby object of interest will be pretty accurate.

 

It's annoying, but works.  Don't hold your breath waiting for a multi-star solution. You could use Cartes du Ciel and Chuck's ASCOM driver. That will allow two-star aligns. But it needs a PC.

I'm happy using a PC, and use Cartes Du Ciel, but I can't see any way for it to do a two star alignment! Can you share how you do this?

[johngwheeler]

If I were doing it, I would use PemPro's drift align, but seeing that the Tak has no PEC, you would get only half the value out of PemPro.

 

Ken, assuming a PAE under 5', how close are the GOTO's?

 

John, do you have SGP? It does sound like closed loop is the way to go.

Yes, I do have SGP, and have used plate solving for centering when imaging. It does take a bit of time though, and I was hoping for a more accurate multi-star go-to alignment without resorting to this.

 

But thanks for the suggestion!

 

John.

[johngwheeler]

it's pretty hard to get sub 5' PA without Polaris visible. You northerners don't know what a good thing you have.

 

i can only get that level of PA if I drift align with PEMPro, which takes forever (well, 20-30 minutes). allegedly the polar scope should work in the southern hemisphere, although I've never seen or tried it.

 

There's an even more idiotic/simple method that I use.

 

Because my EM-11 is a T2Jr (which slews very slowly..) if I have to do a really long slew that requires a meridian flip (I know the slew will be inaccurate anyway and it takes forever) I just declutch and manually move the mount to the other side of the meridian.  Then I center it on a bright star near my object of interest, then resync Sky Safari.

 

Voila.

I've found the polar scope tricky to use well in the south, and it's not helped by having a faulty reticle illuminator (in any case the light would probably be too bright and wash out mag 5.4 sigma Octantis). You have to do an adjustment from the marked time & date to transform it to southern hemisphere values, and even then, it was just a lot of hassle tweaking the mount with the fairly crude adjustment screws. 

 

I have found using SharpCap's plate-solving PA routine to be quite good provided you get close to the pole visually before you start.

 

GoTos are OK of a short distance on the same meridian, but they do get out of the FoV in a C8 if doing a large slew.

 

I'll probably just take a chance and buy Astromist to see if I can get better results with an n-point model. This is just for visual use, and for imaging I'd probably plate solve to center objects.

 

John.

Edited by johngwheeler, 20 October 2016 - 10:39 PM.

[Ken Sturrock]

Ken, assuming a PAE under 5', how close are the GOTO's?

I couldn't say precisely. I'll try to remember to measure it.

[johngwheeler]

Another question I have on the EM-200 alignment is this:

 

The EM200 and Temma driver doesn't appear to offer an index point from which to do an initial slew, like other mounts that have a marked "home" position. Because of this, I've been manually moving to a known star, and then doing my one-star sync on this using SkySafari or Cartes Du Ciel.

 

However, the ASCOM driver settings do have options for defining the "start position" - i.e. OTA pointing West, or counterweight down. I imagine that these are a sort of start position, but because they are rather vague and relatively imprecise (compared to an index mark), I haven't really worked what the point of these is! Can anyone enlighten me?

 

Thanks,

 

John.

[Ken Sturrock]

By default, the Temma will ASSume that the GoTo alignment star is to the east of the meridian (OTA on the west side) and you'll need to pick a star in the east. If you want to align the GoTo on a star to the west (or directly overhead using, say, a bubble level on the lens cap) then you can do that too - but you have to tell the mount.

 

When the mount's expectations aren't met, the mount will be confused and slew the wrong way.

 

If you haven't seen it, start reading at Step 8 on Page 15 of Chuck's manual.

[syscore]

Index marks are used only when the mount has an alignment routine, such as with a Celestron mount. The reason you have to tell the ASCOM driver the orientation of the scope when you sync is because you could be pointing at the star from two different directions (from the east or from the west) and the ASCOM driver has no way of knowing which it is.

 

PS: A celestron mount knows which way it is pointing because you start with it pointing north, and if it moves west then it is pointing west, and if east, then east.

Edited by syscore, 20 October 2016 - 10:53 PM.

[syscore]

I know the Celestron HC pretty well. It probably wouldn't be a huge effort to interface it with an EM-200. The vast majority of its features, including its pointing model, rely only on sending and receiving "mount" coordinates. A small micro controller interface, sitting between the TAK control box and the Celestron HC, could convert between TAK and CGEM mount units and supply the proper TTL serial communication with the HC. The benefit is that the Celestron HC does a pretty good pointing model with its 2+4, and would also provide the ASPA polar alignment routine.

Edited by syscore, 20 October 2016 - 11:00 PM.

[johngwheeler]

By default, the Temma will ASSume that the GoTo alignment star is to the east of the meridian (OTA on the west side) and you'll need to pick a star in the east. If you want to align the GoTo on a star to the west (or directly overhead using, say, a bubble level on the lens cap) then you can do that too - but you have to tell the mount.

 

When the mount's expectations aren't met, the mount will be confused and slew the wrong way.

 

If you haven't seen it, star reading at Step 8 on Page 15 of Chuck's manual.

OK - after reading this and syscore's response below, I understand that simply point at a star doesn't uniquely identify the orientation of the OTA, which might explain why I've had some odd results on occasion!

 

I note that you can start it pointing at the pole (CW bar down or horizontal to east or west), or at the zenith.

 

I'll experiment with these settings and see if this improves matters at all. I'm curious about the Astromist n-point model and it's probably worth the US$23 investment to see if it works as expected.

 

Thanks for the help!

 

John

[orlyandico]

I believe Cartes du Ciel alignment with >1 star is dependent on the ASCOM driver.  I used it with my home-made digital setting circles and the Dave Ek ASCOM driver.

 

So I suspect with Chuck's driver you are out of luck.

 

TheSkyX would be the overkill solution, with the bonus that after building the model it can tell you what your polar alignment is.

 

But for me the entire value of the Tak mount is easy grab and go with only a tablet. Having a PC just makes things unpleasant.

 

If a PC is present, then just use Astrotortilla and plate solve.

[Dwight J]

I have used MaxPoint, The Sky 6, and the Temma ASCOM driver.  I have also had Astromist for a few years and it has been supported with several updates with the most recent within the last year.  I tend to use SkySafari more as the pointing accuracy is good enough for my purposes but I have used Astromist multiple star alignment and it works.  The interface takes getting used to and I haven't used it enough.  Sky Safari kinda spoils you for anything else.  If it had multiple star alignment in the Pro version (hint, hint) it would sell a few more units.   

[Ken Sturrock]

Ken, assuming a PAE under 5', how close are the GOTO's?

I couldn't say precisely. I'll try to remember to measure it.

So, I didn't purposely set out to measure this. I just took the output from the Closed-Loop-Slew routine that ran while I was sleeping last night. As a refresher, what Closed-Loop-Slew does is to slew the mount to the target, snap a picture, plate solve that image, figure out how far away the mount is from the intended target then slew there. For me, the target always ends up dead center on the sensor, after the CLS.

 

The mount was aligned to Polaris with the PAS (took about 45 seconds). I aligned it once and tried to do a good job, but didn't obsess. The GoTo system was aligned to the center of a plate solved image taken at a random location to the east.

 

My first target last night was NGC7331. After the system took the requested images, it slewed to Kochab in the north and parked for a few minutes. The second target that I ran last night was the Crab. So, a little after midnight, the mount finished its rest and slewed from Kochab to M1, which was rising in the east - about 43 degrees or so in altitude. Note that the GoTo had been last (re)synched in the western sky while it was imaging NGC7331. This was the initial M1 slew result:

 

 

 

The grey box represents the border of the plate solved image. You can see that the center of the first Closed-Loop-Slew image is off by a little over 22 arc minutes from M1's actual position. CLS then corrected the position and centered the target:

 

 

After taking 27 images (mostly 10 minute narrow band images) the software decided to flip the mount about 04:50. Previous to flipping, the software re-synchronizes the GoTo system to the target on the east side of the meridian. Due to the way I wrote my software, the mount doesn't flip precisely to the target, it actually flips to a point 1.5 hours west of the target to do a focus, then moves to the target (It's a long story...). Here is the initial CLS image from that point which was 22.5 degrees west of the meridian:

 

 

So, in this case, the slew was off by a little over 23 arc minutes. In my situation, this is no big deal because CLS will fix the pointing error for me. Moreover, even if I was going "manual" - the target would have still landed on my QSI-690's chip and I could have easily framed it. If I were using the same camera on an 8" SCT at native focal length, then the target would have been OFF the chip. As discussed, my polar alignment also benefits from seeing Polaris. So, draw your own conclusions.

 

Tonight, before I kick off imaging (weather gods willing), I'll try to remember to do a horizon-to-horizon slew and see what how much it's off.

[syscore]

Thanks for the measurements. Wow, 22 arc-minutes is quite a bit. I am used to 1 or 2. At least it did get the constellation right. Assuming that your PA was reasonable, and the fact that M1 was in the east and the previous sync was in the west, I would say that this is cone error. You can actually fix cone error the old fashioned way, with a shim.

[Ken Sturrock]

Thanks for the measurements. Wow, 22 arc-minutes is quite a bit. I am used to 1 or 2. At least it did get the constellation right. Assuming that your PA was reasonable, and the fact that M1 was in the east and the previous sync was in the west,

Yes. Enjoy your incoming EM-200.

I would say that this is cone error. You can actually fix cone error the old fashioned way, with a shim.

I know. I'm way too lazy to care. In John's case, though, it might be necessary.

[johngwheeler]

 

Thanks for the measurements. Wow, 22 arc-minutes is quite a bit. I am used to 1 or 2. At least it did get the constellation right. Assuming that your PA was reasonable, and the fact that M1 was in the east and the previous sync was in the west,

Yes. Enjoy your incoming EM-200.

I would say that this is cone error. You can actually fix cone error the old fashioned way, with a shim.

I know. I'm way too lazy to care. In John's case, though, it might be necessary.

 

What is the best way to measure the cone error and fix it? Do the shims go between the RA axis and the saddle, or between the saddle and the OTA? 

[syscore]

There are some methods on the internet. With a fork mount it is easier because it can flip 180 degrees. But with an equatorial, you need to point it at a star overhead, with the weights on the east side, and then do a meridian flip and point it at the same star.  And then (this is my addition) press on the end of the scope and see if it gets closer of further away. I am going to shim for half that distance. I don't know yet what I will use for shim material.

 

There are some other methods that are meant for daylight, but I don't have any decent targets that are far enough away in my backyard.

 

Note: when you do a meridian flip, you have rotated the DEC axis exactly 180 degrees. If the two axes were exactly perpendicular, then it should land right back on the star. But when there is cone error, it lands either high or low, relative to the star.

Edited by syscore, 21 October 2016 - 06:20 PM.

[johngwheeler]

There are some methods on the internet. With a fork mount it is easier because it can flip 180 degrees. But with an equatorial, you need to point it at a star overhead, with the weights on the east side, and then do a meridian flip and point it at the same star.  And then (this is my addition) press on the end of the scope and see if it gets closer of further away. I am going to shim for half that distance. I don't know yet what I will use for shim material.

 

There are some other methods that are meant for daylight, but I don't have any decent targets that are far enough away in my backyard.

 

Note: when you do a meridian flip, you have rotated the DEC axis exactly 180 degrees. If the two axes were exactly perpendicular, then it should land right back on the star. But when there is cone error, it lands either high or low, relative to the star.

Thanks. So would the shim go between the scope and the saddle? I could see all sort of issues trying to adjust the orientation of the saddle on the mount, which would change at different positions. Also, the OTA might not be completely parallel to the saddle, so making the adjustment as close to the scope as possible seems logical.

 

John