10-Micron. Coping with OTA swaps

[CCD Imager]

Hey Ollie

You could put a guider in place when performing unguided imaging and disconnect the relays, then you can have feedback

I may joke, but as I have just started on this unguided lark and after 20+ years of guiding,  I am thinking of doing just that, until I am happy with the reliability of the system.

Adrian

[ollypenrice]

Both systems will work (and sometimes not work!) This is proven fact. Both are, in fact, 'secondary systems' in that they both work on secondary information. In my view the unguided solution works on information which is 'more secondary' than the guided solution.  I'm only challenging the idea that there is something fundamentally more elegant or 'primary' in the unguided solution. To anticipate the counter-argument (as we must!) the seeing does affect the guided solution more than it affects the unguided one, but both are affected.

Olly

[ChrisLX200]

I cannot offer any practical observations not having used mine yet :-) But the GM2000HPS is a very well made mount that happens to track so accurately (thanks to the absolute encoders) that in most cases it just doesn't need guiding. I'm not sure whether 'unguided' tracking was a design requirement or a consequence! Atmospheric refraction corrections are mainly of use below 30deg (as I understand it) and are also an aid to pointing. With a good model it will point anywhere in the sky to about 5 arcseconds so dispenses with the need to plate-solve onto your target. What's not to like? Comparing the current cost against other top notch mounts (like MX+ or AP1200/1600) and the 10-Micron was actually _cheaper_ for me due to the weak Euro/strong Dollar, and I'd rather not have a mount made in the USA thanks - too many issues should servicing be required (the 10-Micron 'brains' are all inside a small box which can be serviced separately). Of course you CAN guide the 10-Micron, and indeed that may be necessary with some OTAs which are not very stable (like SCTs or Newts). A good refractor like the TEC, AP, Televue or whatever is not going to need that however. Guiding is happening 'after the fact' so the error has already happened before you compensate for it, not so with the 10-Micron. The technology is pretty much hidden from the user, it sounds more complex than it really is and creating a model is done during dusk and is retained for next use if your mount is on a permanent pier.

ChrisH

[ollypenrice]

I cannot offer any practical observations not having used mine yet :-) But the GM2000HPS is a very well made mount that happens to track so accurately (thanks to the absolute encoders) that in most cases it just doesn't need guiding. I'm not sure whether 'unguided' tracking was a design requirement or a consequence! Atmospheric refraction corrections are mainly of use below 30deg (as I understand it) and are also an aid to pointing. With a good model it will point anywhere in the sky to about 5 arcseconds so dispenses with the need to plate-solve onto your target. What's not to like? Comparing the current cost against other top notch mounts (like MX+ or AP1200/1600) and the 10-Micron was actually _cheaper_ for me due to the weak Euro/strong Dollar, and I'd rather not have a mount made in the USA thanks - too many issues should servicing be required (the 10-Micron 'brains' are all inside a small box which can be serviced separately). Of course you CAN guide the 10-Micron, and indeed that may be necessary with some OTAs which are not very stable (like SCTs or Newts). A good refractor like the TEC, AP, Televue or whatever is not going to need that however. Guiding is happening 'after the fact' so the error has already happened before you compensate for it, not so with the 10-Micron. The technology is pretty much hidden from the user, it sounds more complex than it really is and creating a model is done during dusk and is retained for next use if your mount is on a permanent pier.

ChrisH

I sense a slight contradiction in terms! If it tracks accurately because of its absolute encoders, what are those encoders doing? Presumably they are correcting for PE and backlash and for changing atmospheric refraction based on a sky model into which the refraction angles have been incorporated. I don't deny that these are excellent mounts and my argument isn't in any way against 10M, it's an argument against the idea that the unguided system is somehow 'purer' or 'cleaner' than the guided solution. In fact both use feedback loops but their source of information is different. Is a 10M mechanically more accurate than a Mesu or whatever other mount you care to name? Do we know? Does it matter? As I said earlier, and as is well known, a mount capable of literally perfect tracking at sidereal rate will not work for our purposes. It needs eiher 'correcting' (10M) or guiding (most mounts.) Correcting and guiding strike me as being not all that different really.

My arguent is only philosophical here!

Olly

[ChrisLX200]

I sense a slight contradiction in terms! If it tracks accurately because of its absolute encoders, what are those encoders doing? Presumably they are correcting for PE and backlash and for changing atmospheric refraction based on a sky model into which the refraction angles have been incorporated. I don't deny that these are excellent mounts and my argument isn't in any way against 10M, it's an argument against the idea that the unguided system is somehow 'purer' or 'cleaner' than the guided solution. In fact both use feedback loops but their source of information is different. Is a 10M mechanically more accurate than a Mesu or whatever other mount you care to name? Do we know? Does it matter? As I said earlier, and as is well known, a mount capable of literally perfect tracking at sidereal rate will not work for our purposes. It needs eiher 'correcting' (10M) or guiding (most mounts.) Correcting and guiding strike me as being not all that different really.

My arguent is only philosophical here!

Olly

:-)  Those encoders divide one complete turn of each axle into 10 million (actually more) increments and that offers an absolute measure of the rotational position. This positional information is fed back to motors such that, when measured against elapsed time, tiny corrections can be made (far faster than following a guidestar) and thus very precise drive rate is achieved. The mechanical accuracy is being augmented by the electronic control directly and not via some estimate as there is with PEC. However, as you rightly point out, having very precise drive rate is not good enough on its own and it needs to be translated into the real world of tracking the stars at varying positions in the sky. That is where the modeling comes in.

If the 'seeing' is less than perfect you are otherwise dependent on the ability of the guidecamera to follow an unstable target, the 10M mount has no such issues, it will follow the true track of where the star should be at any given time - including if the guidestar should dim to the point of being lost (a passing whisp of cirrus perhaps which can cause PHD2 to suddenly jump). Any deterioration of image quality will therefore be directly related to sky conditions alone and not dependant upon the ability of the guiding system to make corrections to drive rate. The mathematics of the sky model is complex but the user doesn't need to know how it works, but there are many variables used (11 or 22 depending on whether the East and West sides of the sky differ - which they may if there is some flexure or imbalance). When applied globally then, corrected tracking is available for the entire visible sky - even in those areas where most of us would not want to image anyway (<30deg above the horizon).

So yes it is both cleaner and purer if you like, the tracking is accurate to very fine limits and in the ideal situation no further corrections are needed. You are not using a surrogate star image - the assumption there being that whatever the guidestar does then so do the contents of the imaging frame (but does it?). The corrections for the 10M mounts on the other hand are wholly internal and can be made at very short intervals - any drive error is corrected almost before it happens.

Chris

I'm sure Per will be along soon to correct me!

[perfrej]

Simple response...

Guiding corrects for imperfections in tracking after they occur but early enough as to not burn the error too deep inro the CCD well. On the plus side, it can handle most imperfections quite well, on the minus side it is a bit slow.

Modeling the 10Micron way corrects for imperfections of the sky by means of assumptions based on previous measurements. It corrects PE automatically as the encoder is in the motor feedback loop. It tracks in both axes, so no need to polar align to redicukous precision. On the plus side, in theory it tracks better than guiding, on the minus is the fact that the tracking is based on mathematical assumptions, not real measurements.

Which is cleaner? Guiding in terms of complexity, modeling in terms of art. I like art

My idea setup, that I am investigating as the ultimate solution, is a 10Micron with a model, augmented with an adaptive optics unit for high speed corrections down to the seeing level.

/per

[Zakalwe]

Interesting as the guide vs unguided topic is, it's drifting a long way from the original point of the thread (ironic given the subject of the thread drift! ).

I appreciate you posting the Google hangout video Per....the Model maker looks impressive. I've spent some time talking with suppliers of the 10 Micron and have come to the conclusion that the 2000HPS is the one that would be perfect for me- lots of capacity and the superb control system. The one ugly fly in the ointment is that I cannot justify £10K given my limited imaging time and crummy location. The 1000HPS would give me superb tracking and pointing, but would require swapping the OTA between applications. From speaking to the supplier he thought that this is an imperfect solution.

I have revised my rather firm rejection of the Mesu as the primary concern was the SciTech control system. I've have spent a lot of time reading the SciTech manuals and trawling the Mesu Yahoo page, and now feel a bit more comfortable with the system. It certainly isn't user friendly or intuitive, but it looks do-able. Plus the Mesu give me massive capacity headroom, which I will probably use by dual mounting my SCT and DSO triplet imaging rig. I will be using it with SGP, so the plate-solving routines will give me the initial pointing accuracy. Plus, there's a handy few £hundred saving between the 1000HPS and Mesu.

[derrickf]

I recently purchased a Mesu200 (not had much chance to use it with work pressures, weather and waiting for a new OTA) and went through pretty much the same thought process as Zakalwe - the GM2000 is the most elegant and technically cohesive solution but it's difficult to justify the cost given the number of nights on which I can image.

Derrick

[CCD Imager]

If you phone Ian King, you might be pleasantly surprised by the current price.....!

[Zakalwe]

If you phone Ian King, you might be pleasantly surprised by the current price.....!

I did...the 2000 is in Paramount MX+ territory. Still too much for me to justify.

[CCD Imager]

My view of the GM2000, is that it is ready for unguided imaging. There is NO periodic error, NO backlash and solidly built. As for expense, well find me a mount with absolute encoders for less than £9000 (GM2000) with similar payload capacity??

The main reason we have had to guide for the last century is due to imperfections in the mount like PE and backlash, however guiding does a very good job and we are too used to it. Parameters such as orthogonality error, refraction index errors plus others are mathematically taken into consideration and there offset calculations are to an order of millionths of an arc sec. As mentioned the mount makes the correction to tracking instantaneously unlike an autoguider which will be a few seconds later depending on exposure time.

However, the major limiting factor for unguided imaging is random movement somewhere in the mount, telescope or CCD camera. My vast experience of 2 nights imaging with the GM2000 (lol) show that not all images have perfect tracking and I need to focus on a couple of areas that need to be tightened down.

Also resultant images depend on the tolerance of the setup, for example a 10M focal length scope might struggle!

As for swapping scopes, it is possible and you should have a saved model for that setup and just re-synch once to update. If it doesn't work, then it is no big deal to wait another 10 minutes for model maker to run a fresh 20 point model then you are back in business.

Adrian

[perfrej]

Adrian, before you start welding all your stuff together you might want to consider running a 50-75 point model and see how it fares. Sometimes that is all that is needed.

/per