Direct Drive Mounts

[Roger273]

Hi All
I'm looking at replacing my EM10 mount, and would be interested in any feedback on the ASA DDM60 direct drive mount. ASA claim you can run unguided with their high resolution digital encoders, but is anyone on iceinspace doing this with the DDM60? It's a similar price to the Paramount MX so any advice on which way to go is much appreciated.
Thank you
Roger

[Peter Ward]

Digital shaft encoders really do not alleviate the need for guiding, as stars often show movement in Dec due standing atmospheric waves.

Drift from the refracted pole will vary from night to night with atmospheric density height, and I haven't seen a in imaging system yet that is immune to small focal plane shifts due gravitational bending moments changing with the camera's orientation.

I've also been intrigued to know what happens to a direct drive mount when the power fails?? With no gear to lock it in place I'd imagine things would get very interesting.

I'm sure they are a fine mount, but direct drive is not a feature I'd rate as essential.

[multiweb]

Spring loaded clutch held by a solenoid? :question: Big clap in the dark.

[LewisM]

Ceftriaxone will help clear that up :D :) :scared2::eek:

[rat156]

Hi Roger,

Again, I don't have a DDM (but I do have an ASA scope), but once the mount is tracking to within an arc second or two, the atmosphere takes over. Ask any imager, there are some nights where you just can't seem to make things work, then you actually look at the stars and see the wild scintillation characteristic of really bad seeing. No mount can fix that (at least in the amateur sphere).

So your question is, do I stump for the DDM and hope I don't have to guide, or do I get a PMX or AP mount and live with the guiding?

I have a PMX, and apart from some problems which did go on for quite a while, it has been great, once I replaced a dud motor. You can PEC train these things to very small residual errors and set up sky models to remove most of the pointing errors, but still most imagers guide anyway. Once you have made that decision, i.e. to guide, then PEC isn't really necessary, a decent polar alignment is all that's required. Even better at long focal lengths is to use AO and only bump the mount when required, if you PA is good, that won't be very often. One of the drawbacks of using a Newtonian scope like the ASA N12 is that you have very limited back focus from the corrector, so no AO for me :(

SO direct drive or not, probably not the question you need to ask. What you get with your mount is also worth consideration. I'm not sure what you get with the ASA mount, but with a PMX you get a lot of software as well. Then there's support, the SB guys run this forum thing, which is frustrating when you have a problem, not sure about ASA, though they have a rep in Melbourne, but I suppose Bintel or Peter are the SB reps in Aus.

HTH
Stuart

[Logieberra]

The new MX+ looks exciting. Can't go wrong with any of the high end guys - AP, SB and ASA (in no particular order! )

[Geoff45]

It seems to me that the direct drive is not going to give you anything better than what you would get with a decent mount (AP, Paramount).
Geoff

[Roger273]

Hi
The feeling I'm getting from Peter, Stuart and Geoff is that the PMX is the way to go, and to forget about direct drive. I'll be in Melbourne next week so I will go and talk to the ASA guy anyway. If I change my mind and go with the DDM60 I will put my experience up on the forum.

Thanks for all the feedback
Roger

[LewisM]

I would read independent reviews on the DDM before the retailer here.

There is a Tak NJP in the trader now at a great price, and these track as well as an AP, PMX/PME/ASA, at half the cost

[Shiraz]

I think that one advantage of the direct drive mounts with encoders is that the main feedback loop runs at fairly high speed, so they can react quickly to control the effects of wind gusts - regardless of whether or not they are being guided. A conventional mount does not have an equivalent feedback loop, so you must rely on mount stiffness (or an AO) to deal with wind.

[clive milne]

Yes, that is basically my understanding as well.
There are a few qualifiers however. Irrespective of the power and accuracy of the drive/ encoder assembly, the maximum useful guiding frequency will be a function of the resonant frequency of the telescope structure and the tuning parameters (proportional, integral & derivative) applied in the control algorithm. It is likely to be the case that the factory default settings will be different from optimum settings derived by experiment in the field.

Also, it needs to be said that direct drives are (in theory) able to manage deviations due to such things as wind loading (particularly in the dec axis) at a frequency and accuracy that is simply not possible using a conventional gear train.

[Peter Ward]

I have no doubt direct drive mounts can give you an near enough perfect rotational rate.

The simple problem remains: that is different to a star's real position due a myriad of atmospheric effects.

Being an "open loop" system the mount simply can't compensate in the same way as an autoguider (or autoguider+AO ) .... which close the loop.

[clive milne]

That is true Peter,
It is also true that a direct drive with an autoguider (or autoguider+AO) is going to have better performance than a geared drive on the same mount.

(assuming it is engineered and tuned correctly)

[Peter Ward]

Unfortunately I've not had an opportunity to test that thesis. What I have found true, using 5-7hz closed loop sampling, is the star sizes that my system generates are only effected by seeing.

I've certainly pondered: would better results be obtained with a drive system that tracks well below the seeing noise?

Running some back-of-the envelope numbers, I suspect my PME is already there, as being in a dome, there is no wind buffeting. Hence, over a 1 second interval, the drive would only have 2arc sec/720 seconds or a .003 arc sec error.

[AndrewJ]

I always thought the big benefit of direct drive is no backlash/friction etc.
DEC would certainly benefit from No backlash,
but without the HiPrec encoder it is ( relatively ) useless for absolute precision/tracking at the arcsec level.
In RA, i would think that ( as the motor doesnt need to reverse ), tracking would merely be a function of the speed of the encoder/motor feedback loop and the accuracy of the encoder itself, incl any reading artifacts.
There has been a very good ( long ) thread on CN recently re the new encoder put out by IOptron for the CEM mount.
The technical/mechanical problems associated with the encoder itself ( when looking at the arcsec level ) shows that a lot of the real gains/losses can come from the encoder itself, vs the drive mechanism.
Very interesting as a system integration exercise.

Andrew

[Shiraz]

I think that there is a bit of a problem comparing mounts, in that maybe we don't know for certain how our mounts actually perform.

We might try to find out by tracking a star for example. We get data that shows the star wobbling about and immediately say "that high frequency stuff is seeing noise" - what is left after the wobbly bit is filtered out must be the mount error. But maybe part of the wobbly bit is also mount error - we simply do not know. It would not be surprising to find errors due to minor gear and bearing vibrations, or noise from dirt in the lubrication etc (1/100 the thickness of human hair will give you an arcsec of error). In principle, a direct drive mount could have lower levels of chatter if properly set up - but who knows.

Anyone know of any direct angle/rate measurements for a mount, as opposed to star position?

WRT your original question Roger, apart from ASA, quite a few high end systems use direct drive, so it clearly is worthwhile - eg: http://planewave.com/technology/mechanical-design/
http://www.skyvision.fr/wordpress/mount/direct-drive-mount/?lang=en
http://www.astelco.com/html/products/ntm/ntm.htm

[DaveNZ]

Interesting video which talks about Planewave direct drive about have way through. https://m.youtube.com/watch?v=B_Louir2r10

[marc4darkskies]

No doubt, direct drive would definitely be a nice-to-have. My biggest problem though is that no-one seems to have engineered a solution to the problem of unwanted movement when the power is off. This seems like a simple thing to solve and I'm puzzled why ASA doesn't do anything about it or even acknowledge or address the question :shrug:. They seem to rely instead on a perfectly balanced system ... and I've never seen one of those - especially when you have a rotating camera whose load is not perfectly concentric with the optical axis! Also, what happens if power is lost while slewing - an admittedly improbable event, but a very ugly one. If I'm wrong, someone please tell me because that's probably the only reason I would never consider buying one.

Cheers, Marcus

[AndrewJ]

Gday Marcus

"Most" precision direct drive motors are to all intents and purposes just bloody big steppers with advanced microstepping mechanisms to control speed.
Depending on the design, the motors used in the ASA may actually have a holding torque ( like a std stepper ) that defaults to keep it stable when the coils are no longer energised??? Dunno, but if anyone has an ASA mount, maybe they could advise what happens if deenergised???
Ie if the clutches are left locked, and the mount isnt powered,
how much torque is reqd to make the axis rotate ( if at all )?

Andrew

[Shiraz]

Maybe use a UPS when likely to be active and manual lock when not. Then, provided that all plugs and cables are high quality, the only possible failure would be if the drive circuitry went bad. But we accept the very slight risk that the electronics of our worm drive machines could malfunction and slew the mount into the limits, so failure of drive electronics is probably an acceptable risk for a direct drive as well.

However, it would be worth ASA at least addressing the issue rather than pretending they don't understand the question on their FAQ page. Otherwise they are going to get killed in the marketing wars.

[marc4darkskies]

Here is a very interesting post in an ASA forum about a member starting up a DDM in a remote observatory after power off:

"... Mount is switched off and parked, As it is equilibrated with the telescope and CCD ...mount keeps perfectly parked.. When I want to begin a session..I open the roof... switch on remotely the mount and by means o Teamviewer software.. looking through a webcam very sensitive sited inside the observatory.. I use Autoslew in order to move the mount aproximately to the zenith. Then I click Homefind and in few seconds the mount finds the exact values. Then I open a planetary programm (The SKy6) and synchronize with a near star. From this moment..I can begin any astronomical session automatically ( Used ACP and CCDAutopilot)"

The bold text is mine. Summary points seem to be:


Needs to be balanced to maintain a parked position!
When starting up from power off you need to point to the zenith first (using a remote camera)?!
Then and only then can you tell the mount to find the home position!
Then you have to sync on a star before starting your session!

The poster then goes on to say this always works and he doesn't understand where the problem is!

I'll tell him where the problem is! Paramount: 1) switch on 2) tell the mount to home (from any parked position) and start a session. And when the power is off the Paramount won't move - guaranteed!

[Terry B]

And for my Tak NJP it is similar. Turn the power on. Unpark the scope. slew to any star- usually within 20 arcmins. Take an image and platesolve to fine tune the pointing then slew anywhere.
Fine balancing is impossible due to the odd shaped spectrograph hanging off the end of the scope. Doesn't seem to matter though. The motors are powerful enough to cope.

[Bassnut]

Oh dear, if all this true, that's just woefull :shrug:. No wonder you don't hear a lot about them. The price isn't bad though, for direct drive, but rest is just trash. Itelescope tell me the CDK direct drive is the ducks guts and very reliable. Albeit far more expensive.

[clive milne]

Actually, I would suggest that the mount should be imbalanced such that on loss of power (and no breaking) the scope will automatically settle in a predetermined orientation. ie) If you made the counterweight shaft heavy on the RA and primary mirror end heavy on the DEC, the scope will automatically park upright and pointing at the pole. You could have the imbalance torque 10% of the motor torque to achieve this without any meaningful loss of performance.

I would also suggest that the shortcomings of the ASA system are not necessarily inherent to all direct drive telescope mounts.

[Roger273]

Is everyone just assuming that loss of power is a problem with the ASA? I don't think this thread has any comments from someone who has actually used one as yet. The following email is from a user who sounds more than happy with his ASA:

"Dear Roger,

My relationship with ASA dates back to 2006 as they adopted my proposal to produce Astrographs with fast focal ratio. Since then I have tracked the developments of all their projects. One of the highlights was, as they passed to me a DDM85 direct drive mount. Previously I gained my experience with different traditionally mounts (GPDX Vixen, Losmandy G11, Losmandy HGM Titan, AP 1200). But with the DDM 85 from the very beginning I was sure that I never want go back to a traditionally mount. I assembled off axis guider systems at my observatories but most of my images are taken without external guiding. My portable mount is an ASA direct drive (DDM 60) and does the same good job as the DDM 85. I´m really very happy with both mounts. As I know ASA has an Australian in Victoria – have a look at their website".

Other comments from anyone who has used an ASA mount would be greatly appreciated.
Roger

[Bassnut]

ASA product support generally is another matter that needs considering. At least one high profile user on IIS and another serious user in OZ im aware of that wasnt public with his experience had the same absolutely jaw droppingly bad experiences with ASA engineering and support. I mean expensive product engineered so badly it was totally unusable, and next to zero support. Things may have improved lately, but I wouldnt go near them myself, very risky. I

[Shiraz]

Interesting question, so asked ASA what protection there is against mount failure with power loss. Their response in summary was:
- a reasonably balanced mount will just stop when power goes off - due to bearing friction.
- the mount software detects if the scope contacts anything when under power (position is not as expected) and powers down - unlike a geared mount which will keep trying to drive.
- the mount software can do a "homefind" on power up.
- there are more than 300 DDM out there and they know of no examples of scope damage due to power outage.
- they haven't fitted a brake because there is no need for one.

FWIW, have confirmed that my EQ6 and EQ8 both stop dead and hold position if the clutches are released, even if slewing (ie as would happen with power out on a DDM). Provided that the loads are fairly well balanced, some extra torque is required to overcome bearing stiction, as ASA state. Of course, the main servo loop in a DDM will need to be tuned to the load, so good balance is essential anyway. The idea that a DDM will somehow fall over or continue slewing if power fails seems to be just another urban myth - plausible, but not real.

Looks to me like the DDMs could be far more robust (in principle) than has been suggested - they just require a slightly different set of procedures and assumptions.

Edit: just looked up the homefind process - the mount does not need to be manually positioned pointing to the zenith to initialise, it's just that this is where it will most quickly find the reference marks that tell it exactly where it is.

[frolinmod]

That may be true for stepper motor drive systems, but I'm not aware of any servomotor drive systems that keep trying to drive. So far as I am aware, when they hit something they all detect the excess motor drive current and immediately stop dead in their tracks. Paramounts also sound the claxon after stopping, just to embarrass and humiliate you further. :-)

[AndrewJ]

Gday Frolinmod

Not sure re the Meade Classics, but none of the later Meades "measure" current for this purpose. ( Hence why motor cards can burn out )
They monitor the encoder and if the motorcard has been told to move the motor and the card needs to use a duty cycle that exceeds a set value to try and move the axis and still NO movement is detected over about 16 seconds, it declares a motor fault.

Andrew

[GeorgeC]

This thread is a bit old, but I only just stumbled across it.
I own a DDM60Pro mount. Regarding the balance question... Yes, the mount must be very accurately balanced to get good performance. I use a system of sliding weights to get near perfect balance.
If there is a power cut (not happened yet!) then there would only be a problem if it happened during a fast slew. If it happens during tracking the residual friction in the bearings will keep it in a fixed position.
The issue of unguided imaging is vexatious. If your scope has ZERO flexure, in all planes, in every position, then it is certainly possible. Some people get 45 minute unguided images.
Even with a sturdy carbon fibre tube I can not achieve it and so guide.

Older DDM mounts did have to be calibrated at the start, but ASA now put absolute encoders in, and the mount knows where it is when switched on.
Some time later this year there will be Version 2 of the mount which promises to be even better (but probably a bit pricey!)

[Peter Ward]

After talking at length about the subject of direct drive mounts with a well known Californian mount manufacturer, seems the elephant in the room is the rigidity of the mount in the first place.

Big mounts have often have big gears for a reason.

The simple fact is a large diameter gear will hold things in place with significantly more rigidity than a magnetic field.....and without deflection compensation, direct drives would be pretty "rubbery" indeed.

Sure gears also have errors....but these reduce through simple geometry with gear size (assuming the gear is equally well made) and absolutely encoded geared mounts are easily as capable in being sampled many times per second to effect a perfect tracking rate with far less susceptibility to be deflected by buffeting in the first place.

Wind buffeting on the OTA may or may not be sensed as a rotational rate change (mechanical stiffness of the OTA may not be great) hence its a bit of a red herring to say it can be compensated for all the time.

[Shiraz]

'fraid that simply doesn't match with what is going on everywhere else in astronomy Peter. have a look around the web - every mount of larger size uses either direct motor drive or friction drive. Starting from the top-end Planewaves and all the way up to the huge VLT class scopes (the VLT has a 10m diameter direct drive az motor), it looks like nobody uses gears anymore. The primary advantages given for direct drives are rigidity, precision and resistance to wind buffeting. They are not in any way "rubbery" and having rapid feedback to respond (without gear lag) to wind deflection is a big plus.

That is not to say that cost-effective gear drives do not have a place in our small, low-cost amateur systems, but I doubt that there are any technical reasons why direct drive systems with absolute encoders would not do a better job - regardless of what a maker of good quality gear drive mounts might say.

[GeorgeC]

You can tune the motor parameters to make the mount response 'stiff' or a bit 'rubbery' if you want. There is a critical point where if it is too stiff you can get oscillations, so most times you tune it to that point and then soften it a bit. The response to wind is amazing - if you wave a piece of A4 paper near the scope you can see the mount current immediately rise to correct the deviation.

[Peter Ward]

The decision to go with direct drives on extremely large mounts is cost driven. It is simply not practical or cost effective to make massive gears for the likes of the VLT.

Large diameter high precision gears are really expensive, and difficult to make well, which I'd suggest is the real motivation toward direct drive systems in larger amateur systems.

A fully meshed gear also has no lag. Backlash is a hallmark of cheaper systems and you should not see any in high end mounts by Astro-physics, Bisque etc.

Some mount manufacturers also accelerate the drive motor for a short time
if they reverse direction. This removes any less than perfect meshing delay...which is almost exclusively on the Dec gear, as the RA drive only needs a small rate change which can applied virtually instantaneously.

As for either system doing a better job, with absolute encoders fitted on each, the point is moot. Both are seeing limited. Both respond well to any sensed rate change from wind buffeting....and I still maintain a gear will hold a telescope more rigidly than a magnetic field.

All that said, having used a direct drive mount recently, I can say positively tracking simply isn't perfect. If you don't guide, expect eggy stars....particularly if you take 20-30 minute subs as I often do.

Their albeit perfect drive rates are open loop systems that have no way of compensating for atmospherically induced drift in RA and Dec....hence I only see "advantages" of direct drives as marketing hype rather that offering a practical difference.....unless of course you are running the VLT :)

[GeorgeC]

If wind gusts hit a direct drive mount in the DEC axis direction the response is almost instantaneous, and has the same accuracy whether it is compensating North or South. There is bound to be a small oscillation so the correction impulses will alternate rapidly between N and S.
Even with the most precisely engineered geared system, I think it is asking a lot for the stepper motor to give a series of corrections in alternate directions without any time delays as the chain of gears do their work.

I may be wrong...

[Peter Ward]

Two things. My PMEII is in a dome, so buffeting by wind gusts is as likely as that by pink elephants. A perturbed high stiffness structure will likely damp out
more quickly than one which has to be actively modulated to achieve the same effect, as like a plain old spring, it intrinsically wants to snap back to its unperturbed state.