TDM vs Autoguider?

[originaltrilogy]

I have CGEM about 2 years old.

Conntemplating getting a TDM Telescope Drive Master. It says it can reduce PE to <5

Question is, what benefit is TDM over autoguider? Dim targets maybe?

Would TDM really make CGEM as good as Mach1 for example?

[Poita]

Lots of views and no responses, so I'll have a crack.
The TDM system adds a very high resolution encoder to the centre of the mount itself (replaces the polar scope I think).
This allows the system to monitor where the mount is *actually* pointing, rather than where it thinks it is pointing.

So corrections, very accurate ones, can be applied very quickly, which would remove things like PE and even wind gusts and bumps I would imagine.

So assuming your polar alignment was pretty much perfect, you could get away without guiding for quite reasonable length exposures.
No guidescope or guidescope flex to worry about, one less camera, no issues with having to find a guidestar, accidentally guiding on a hot pixel etc.

Downside is that if the seeing is all over the place, it won't help you, or if your polar alignment is off.

Apparently the new model allows a guidecam as well.

I would think it would work great with an adaptive optics setup, the TDM would make sure the scope was tracking perfectly, with almost no PE etc. and the AO would take care of refraction/seeing/slight abberations and so on.

It is pricey though, about $2000, but if you already have an old EQ6 or CGEM, then you are payinh $2000 to get a mount that has a PE of potentially < 2" which would mean all you have to worry about is your polar alignment.

It is transferrable between different mounts by utilising different adapters, so there is some potential for changing around if you get a different mount.

I would think it should be possible to 'push' the mount around and not have it lose it's marbles, as the TDM would know how far you moved it.


Downside:
PA has to be very good, no accounting for atmospherics (though if the seeing is that bad, do you really want to image anyway?)

[Poita]

As to whether it will be as good as a Mach1, well... PE wise, I would say yes, carrying capacity, I don't think so.

[wasyoungonce]

Well I'll mention some other ideas, FWIW.

Atmospheric seeing conditions are around @2~3 arc sec's at best (guesstimate avg)...so you cannot RA track/guide better than this. By track...lest say if you are RA tracking at perfect sidereal rate...the stars will still oscillate around regardless.

TDMs are expensive...auto-guiding is cheapish and can attain good results with guiding. I'm not sure what CGEM users get but lest say around +/- 2 arc seconds with PHD. :shrug:Seeing will be on top of this as it would with a TDM.

TDM works in RA only...what if the polar alignment is "off"?

IMHO...the best and far cheapest option is off axis auto-guiding.

[Poita]

I guess if you are autoguiding as well, then the autoguider has less corrections to do, the mount should already be tracking solidly with the TDM and the corrections made very quickly so you end up with very smooth performance, no sudden movements should mean tighter images.
Or if using adaptive optics, then you are cleaning up the seeing, as any adjustments required would then easily be within the AO capability.

Just like the Ovision worm for G11 mounts, you want the base performance of the mount to be as smooth and error free as possible, the TDM might deliver this performance from cheaper mounts, saving you the hassle of selling and buying a more expensive mount. You can keep what you have already and improve on it.

I'd jump on it if it was a $500-$1000 add-on, at $2000+ it gets expensive.

[gregbradley]

I know nothing about the TDM but there is more to PE than numbers. Its also how smooth the PE and whether there are sudden spikes or bunps.

I'd be very surprised if any addon could match an AP Mach 1. Sounds like wishful thinking.

Also a Mach 1 with Pempro would most likely be way below 5 arc sec PE and the PE would be smooth not rough.

Greg.

[marki]

Greg the TDM is a very high quality encoder which is bolted to the RA axis and attached to a control system. This gives very accurate readings of the absolute RA position and movement faster/slower then sidereal rate is removed via control of motor speed. It does not matter too much whether the PE is uneven or spiked as the encoder updates the control system many times per second which then corrects for errors. Basically a mechanical/software solution for average to poor worms. In theory you could use such a system to flatten the PE curve to AP like precision, in the real world hmmm. I like my EM200, basic as a farm tractor but very precise worms :)

Mark

[gregbradley]

There is always a mechanical lag on corrections, the motors have to move considerable mass. So rough worms may be tough to correct for completely because of that. I have seen a site that coated worms of EQ6 type mounts with some sort of super lubricant that claimed big improvements in PE.

Perhaps using both would get good results - the mechanical smoothness needed plus the accurate encoders.

Greg.

[Poita]

The PE becomes very smooth and almost non-existant if you look at the graphs for the TDM on an EQ6.

I would say if you were going to add a TDM then yeah, it would make sense to give the mount a go-over and clean up and lube the gears well. The TDM can also be added to CGEM, G11 etc.

[Poita]

There is a similar approach and write-up here that may answer some of Petr's questions.

http://www.cloudynights.com/item.php?item_id=2750

[gregbradley]

A very interesting link Peter. That seems to be the future of mounts. More accurate encoders are becoming more common as an option for mounts.

Greg.

[AndrewJ]

Gday Greg



I agree that more accurate encoders are the way to go, but you also need to consider where in the design chain they are used.
If i understand it correctly, the TDM is an aftermarket unit that effectively sends precise guide commands to the scope. Thus the system is still at the mercy of how the mount handles guide commands, unless it totally replaces the hosts motor control system.
If a hi prec encoder is designed into a mount from day one, great, but how you meld an "add on" into a variety of "proprietary" scopes may have interesting consequences.
Considering most mounts already have a PEC ability, i always wondered why someone didnt put a "mid precision" encoder onto the worm.
Assuming say a 180 tooth worm wheel, and that you could make a PEC model up, you could get away with a 180x less precise encoder for the same tracking accuracy, and have no need to fix anything to the RA or DEC axles, thus making a much simpler installation.
As the encoder is on the worm, any gearbox PE is eliminated,
and a single turn PE model can be used for the remainder.
Much much cheaper for a "retrofit" type installation.
Thoughts????

Andrew

[marki]

Andrew the TDM by passes the worm completely and is fixed directly to the RA shaft itself. The thing with this is it has no effect on the dec axis so if your polar alignment is out even the smallest bit the whole thing becomes a paper weight.

Greg I believe the future in mounts will be direct drive and most the problems of dealing with worms and gears will become a distant memory only to be replaced by new problems.


Mark

[AndrewJ]

Gday Mark



Understood fully, but all i am saying is, on a cost benefit analysis,
can a much cheaper encoder on a worm + PEC be equivalent to a really expensive encoder on the output axis of an average quality mount???
Any mount designed from the ground up with a direct drive motor and an integrated hi prec encoder "should" always have the abilty to outperform a TDM retrofitted to a crappy mount.
Horses for courses.

Andrew

[marki]

Andrew I don't see how that could improve things as an encoder mounted directly to the axis would give an absolute position (we are here) whilst one mounted on the worm would still be giving "this is where I think we are" and would be at the mercy of the quality of the worm once again.

Mark

[AndrewJ]

Gday Mark



Understood, but some of the high precision encoders arent "absolute" encoders, they just have high precision.
The high precision "absolute" encoders are horrendously expensive, relative to a retrofit unit that is just used to get better tracking.
Again, all i am musing is, could a big gain in accuracy ( relative to present ) be gained by using a suitable encoder on the worm.
This wont remove PE, but it means it can be mapped on a true 1:1 basis, hence removing all the quirks of a gearbox with odd gears.

Andrew

[gregbradley]

Andrew its worth a test and if it works you can sell an "Andrew worm encoder add-on" for $500 and sell them all day.

I read in the Bisque PMX manual that mostly errors in the worm (20 microns) cause the PE. 20 microns is a pretty small size. I assume that figure is larger for less well machined worms.

Greg.

[marki]

Greg, the accuracy of the machining process will be neither here or there these days as the vast majority of worms and worm wheels would be cut on CNC machines with better than 0.01 mm accuracy as it would be the only way to meet mass production quota's. The real difference is made in what takes place after the machining process to clean up the worms and this is usually in the form of lapping with various grades of diamond grits to remove the lumps. It is very time consuming hence the extra cost. Brendan Mitchell spent quite a bit of time lapping his worm gear mesh on a simple EQ6 and gets fantastic PE figures but he has also set the gear mesh clearances' very well to. Another factor is the size and number of teeth on the worm wheel itself, basically bigger is better as small errors will be less pronounce and mount builders such as SB and AP always use large wheels on the RA for that reason. It is best to view the machining process as a raw stage before final finishing. Mass produced mounts such as those made by celestron, meade, skywatcher etc do not have the same attention to lapping as something like a PME/PMX or AP mount but they don't need to as the profit margin is not there. Other things such as worm block design and run out in the worm itself will also come into play but these will affect backlash more then anything else.

Andrew, why not give it a shot on a cheap mount and see what happens. My EM200 has the encoders on the end of the worm so you might be on to something.

Mark

[AndrewJ]

Gday Mark


I only have Meades and an EQ6, so too hard.
The EQ6 doesnt have the space, and with the Meade, i know more than enough about how the firmware works to know that reverse integrating it would be more of a problem.
Getting it to track at a set rate may be OK, but how to control motor reversals ( esp in DEC ) would be more effort than its worth.

And just for info re the manufacturing tolerances, i did the numbers on a meade wormwheel a few years back
With a 5" dia wormwheel, one arcsec of error at the point of contact with the worm is approx 0.00031mm.
Puts it into perspective how small the tolerances are

Andrew

[marki]

Andrew the EQ6 is a Chinese copy of the EM200 so I imagine fitting encoders to that mount would not be too difficult. The encoders on the EM200 are outside of the case directly attached to the worms (see pic below).

Mark

[AndrewJ]

Gday Mark


I assume the EM200 has extended shafts on the worm itself????
My EQ6 has no free section of worm axle to easily hook anything onto, and as the new unit would have to provide feedback via guide commands, i would still need to make a software black box to drive it.

With the Meades, its simpler to hack into the standard processing loops of the firmware, ie semi seamlessly replace the current encoder with the new one, but it still leaves a problem of how to deal with backlash in an "elegant manner". Ive been looking at how that works in the Meades for several years, and it is not a trivial exercise.
I guess if it was simple, it would have been done by now.:thumbsup:

I still like the precision of a direct drive motor with an absolute encoder on the output axle as being pretty close to optimum.
Just hope the prices come down with time and scale of manufacture,
and all this conjecture will become moot.

Andrew

[marki]

Yes the worms are slightly longer but the encoders slip inside the worm and are fixed with a grub screw. Would it be possible to drill a hole in the centre of the EQ6 worms and do the same? Most likely I would think, wouldn't prove too difficult. Now the black box, well that's a different story unless you worked out a way to wire them back into the circuit board internally and made corrections in the software to compensate for the slower rotation.

Yes direct drive is the future no doubt.

Mark

[AndrewJ]

Gday Mark



Possibly, but to get good precision of the bored hole would require
an accurate setup, plus modifying the preload endplugs.
Ie probably not worth the effort on the EQ6.

Andrew

[Poita]

I'm having a brainfart, how would the encoder on the worm work?

[AndrewJ]

Gday Peter



Same as per usual, but the position of the encoder determines how much "guessing" the system has to do.
There are also 2 scenarios here
a) Use an external Hi accuracy encoder to give tracking corrections via the guider port.
b) Use a mid accuracy encoder on the worm to directly replace the feed from the existing encoder on the motor.

I am familiar with the Meades ( which already use encoders ) so would choose option b) and explain it the following way.
With an LX200, the encoder is mounted to the motor.
There is then a 50:1 gearbox then a 180:1 worm drive.
Assuming we had an absolute encoder on the final axis,
to get say 1 arcsec resolution, it needs 1,296,000 counts per rev
If we put the encoder on the worm, it needs 7200 counts per rev
If its on the motor, we only need 144 counts per rev.
If you put the encoder on the output axle, then if it moves, the axis is moving, ie there are no backlash or PE errors between the encoder and the OTA.
If its on the worm, ( and the worm is springloaded so always in contact ), there is virtually no backlash, but there is still the worms PE.
If its on the motor, you have all the lash and PE in the gearbox plus the PE in the worm.
When tracking, the system needs to know the PE, and when guiding ( esp in DEC ), the lash needs to be constant ( and its not )
The Meade LX200 gearbox requires three revs of the output axle before all the gears resynchronise ( hence the three turn PEC model they use )
The LX90s etc require 128 turns or the worm for the gearbox to resynch.
As such "modelling" PEC when it includes a gearbox like this is a non trivial problem, esp if the gearbox PE is large relative to the worms PE.
An encoder on the output axle is optimum, but horribly expensive ( at present ).
A mid range encoder on the worm is a much cheaper proposition, but still needs a PEC model to be used to get good tracking. However, as the encoder is on the worm now, no PE effects from the gearbox contaminate the model, hence an accurate model can be generated.

So its just looking at it on a cost benefit analysis of what method gives the best bang for the buck.

Andrew

[marki]

Andrew I think it would actually be pretty straight forward mechanically. Simply drill a 3.5 mm hole in the desired end and thread m4 then make up a sleeve to suit the encoder shaft with a m4 male thread on one end. You would not need to alter the worm block at all. Now the software that's another matter entirely.

Mark

[AndrewJ]

Gday Mark



Sounds simple in practice, but to drill "a perfectly concentric" hole into the end of an existing worm, without any runout is not a simple exercise. And i make this comment based on the tolerances we are talking about.
At the precision we are talking about, getting the encoder mounted without any eccentricity is a critical requirement to getting repeatable results.
My lathe, even with a collet fixing probably isnt up to this :)

Andrew

[marki]

Andrew, the tolerance you have stated is fine for the worm and worm wheel but are not necessary when mounting encoders. For example when I run the EM200 at full slew you can plainly see the encoders oscillating around the central axis as a result of them being fixed to the worm by grub screws. My EM 200 delivers +/- 2arc sec PE without PEC (you don't have a choice on the PEC, there isn't any nor does it need it) and when properly polar aligned drops the target smack bang in the middle. I build miniature engines with my lathe and mill and my latest project, a V8 required me to drill the camshaft hole 100mm long with a max tolerance of 1 thou run out over the length. With careful setup and patient drilling with a sharp drill and lots of coolant I got the job done. Mounting an encoder would not require anywhere near this accuracy and the hole would at most only need to be 5mm deep. I base this statement on using similar encoders to those found on the EM200 which are basically glorified volume pots. You may have a different system in mind.


Mark