Perfect drive system in development

[Satchmo]

Sidereal Technology have developed a drive control for new style direct drive system for Telescope mounts.

The advantages for telescope control are :

1. Completely eliminate all backlash
2. Completely eliminate all periodic error
3. Completely eliminate all non-periodic error
4. It makes a fast, responsive system, able to counteract small wind gusts
5. Fast, smooth slew speeds, up to 45 deg's per second

Read all about it :

http://www.siderealtechnology.com/DirectDriveWorkInProgress.pdf

Great to see some new innovation.

[davewaldo]

Wow! this looks very promising! Looks like it will be affordable too!

I hope they can create a commercial product before someone copies them.

[h0ughy]

wow that is impressive - but now way would i put a computer near it, or a memory stick, or for that matter a camera!!! those magnets are very very powerful!!!!!

[tornado33]

Interesting idia. 10 ticks per arc second is very good resolution. I wonder what torque the motor can do.
Its a wonder manufacturers dont want to pick up on this idia.

Years ago washing machines went direct drive, its amazing that now we may see direct drive motors with a 1 rev/24 hour speed. If something jams, stalling it wont hurt anything either I imagine.
Scott

[bojan]

At some stage I was wondering about Ultrasonic motor, used in Canon lenses..
All in all, this looks very much like a conventional stepper motor, microstep driven. Or, like the (flat) motor used for floppy-drives.
The encoder used is very expensive part, though...
Also, the trick is to provide enough torque, without gear reductor. All periodic and non periodic errors can be eliminated here because it is closed loop (so the only critical part here is encoder, it must be accurate).
Very interesting.. the question also is, would it be cheap enough to manufacture? Because if it is not, it will not sell... and it will disappear as just one more interesting idea.

[Bassnut]

They cunningly chose to exclude the encoder from the cost guess in that PDF. I went to an industrial show where Renshaw had a stand and had a look at the encoder set these guys are using. They fitted an encoder set to a pro scope in Oz while they were here, and these have been used on a few scopes around the world. The cost of ONE encoder set my friends, is $7000 !!!, thats $14000 for a pair (although possibly not necessary for DEC).

Direct drives with output encoders is not a new idea, its the obvious optimum solution, and has been common in Industrial Automation for a long time.

What keeps it away from amatuer astronomy, is the encoder cost. Output encoders at the resolution required for Astrophotography are quite new and increadibly difficult to manufacture.

You wont see Direct drive/output encoder mounts for the amature market (at a reasonable price anyway, say at or less than a PME) anytime soon, unless there is a revolution in encoder design.

[avandonk]

That is the heart of the problem very accurate encoders. This can be simply done ....I will not tell yet.

Bert

[Bassnut]

Actually Bert, its not so much the accurracy, its the resolution thats the problem, accurracy can be counted with a compensation table. The Renshaw guy said calibration is also a huge problem, what is the reference when the encoder to be calibrated is the highest resolution available???. And, as a result, they prefer to calibrate after installation, also the physical properties of the machine it is fitted to almost entirely determines the accurracy, very expensive.

Can be simply done Bert, pray tell , you would be a very wealthy man if you could do that :-).

[Satchmo]

I'll email Dan and see what he says about the encoders. It doesn't make sense that he would get so enthusuistic about the system for ATMers if the encoders were going to be $14k .

[bojan]

Resolution, according to pdf file is adequate (10 ticks per arcsec).
But the price is definitely not.
My money is on conventional systems. The reasonable thing is to have maximum bang for a buck..
Unless someone just MUST have something like this.. like if someone wants to drive Ferrari instead of Toyota.. and can afford it.

[avandonk]

Bassnut it is not only the encoders but even on an EQ6 say encoders that have been around a full 360 degrees and all the while by software logging the 'errors'. You do not need expensive hi res encoders just very smart software and hardware. The system can be almost totally predictive of upcoming errors.
Bert

[Bassnut]

Bert

All amature mounts have encoders mounted on the drive motor, before the worm drive, so are not the true closed loop output encoders offer. Motor encoders are very cheap and low resolution (but mutiplied by the gear ratio), but the mechanicals after that are not in the control loop so their precision directly affects tracking accuracy. Active guiding is a slow, reactive, iffy compensation. Thats the key, all though PEC is somewhat predictive (and non periodic aint), guiding is reactive, not predictive.

Output encoders provide the ultimate closed loop control, so as long as there is no backlash (as in a direct drive, no gears), the mechanicals are then irrelivant.

Yes, PEC on motor encoder mounts is effective, but no where near the closed loop accuracy provided by output encoders. In fact, theoretically, guiding is redundent on a true fully closed loop system.

[Wavytone]

Not entirely new - the same style of motor was commonly used in some floppy and hard disk drives decades ago, especially with floppies as the windings could be assembled on the printed circuit card.

Another bit of technology some of you might find interesting is called the "harmonic drive". This is a gearbox with zero - and I really mean it - zero backlash.

At the moment harmonic drives are very hard to source (UK and Germany) and cost a lot however some miniature ones have been made.

I can't help feeling a direct drive motor and a harmonic gearbox would be an awesome combination.