What can I use this encoder for?

[Ian PParticipant @ianp]

I have had this lying around for a five years now and have yet to find a use for it!

Its a programmable absolute encoder which gives 4096 pulses per revolution, and, counts to 4096 rotations. If I fitted a rubber tired wheel with a circumference of 1m I could accurately measure distances up to 4km. Thing is, I don't think I will ever need to do that!

Anybody got any suggestions as to what practical purpose I could put it to in the workshop

[Ian PParticipant @ianp]

Has 0.25mm resolution over 4km (with 300mm wheel)

[John HaineParticipant @johnhaine32865]

A GREAT manual pulse generator for your CNC mill or lathe!

[John Stevenson 1Participant @johnstevenson1]

Electronic gear hobber.

See MEW 108

[Michael GilliganParticipant @michaelgilligan61133]

Very high-resolution Indexing for a Rotary Table or Dividing Head ?

MichaelG.

.

P.S. if you need the Manual, this appears to be it.

Edited By Michael Gilligan on 03/04/2014 22:40:22

[Les Jones 1Participant @lesjones1]

Hi Ian,
You could connect it to a leadscrew on a machine to give a position readout. (I used this method for my first attempt at fitting a DRO to my milling machine.) You could also use it coupled to s DC motor together with the "UHU Servo" design. This would make it behave like a stepper motor but with closed loop feedback.

Les.

[Ian PParticipant @ianp]

Posted by John Haine on 03/04/2014 22:38:10:

A GREAT manual pulse generator for your CNC mill or lathe!

John

I have seen the wheels/knobs fitted to CNC machines but never thought there was anything special behind them. I assumed there would be a bit of a flywheel and an encoder. I would not think that an encoder with a large number increments would be needed as presumably there is a layer of software involved to get the right 'feel' to the system.

I think the encoder I have would be wrong for MPG, and incremental type would be more appropriate.

Ian P

[Ian PParticipant @ianp]

Posted by Michael Gilligan on 03/04/2014 22:39:34:

Very high-resolution Indexing for a Rotary Table or Dividing Head ?

MichaelG.

.

P.S. if you need the Manual, this appears to be it.

Edited By Michael Gilligan on 03/04/2014 22:40:22

Michael

I have some of the documentation for the encoder but had not seen the one you linked to, so thanks. As you suggest it would make a very high resolution indexer, coupling it to a rotary table is the tricky bit though and any gearing is likely to degrade the potential accuracy. The fact that it is an absolute device would not really be any advantage and because its output signals is serial (RS422) it needs a PC or microcontroller so starts getting a bit involved (for me anyway).

Ian P

[Ian PParticipant @ianp]

Posted by Les Jones 1 on 03/04/2014 22:59:34:

Hi Ian,
You could connect it to a leadscrew on a machine to give a position readout. (I used this method for my first attempt at fitting a DRO to my milling machine.) You could also use it coupled to s DC motor together with the "UHU Servo" design. This would make it behave like a stepper motor but with closed loop feedback.

Les.

I think that rather than coupling it to the leadscrew it would be better to fit a small drum to the shaft and convert it to a 'wire encoder'. Keeping it free of swarf is the main problem.

Ian P

[Neil WyattModerator @neilwyatt]

If you had two, you could combine them with friction clutches and stepper motors to make a 'go to' telescope mount that can also allow for any manual shifts you make, although the resolution is a bit low for actual equatorial tracking.

Neil

[Ian PParticipant @ianp]

Neil

I don't know much about telescopes and astronomy but if a 4096 increment encoder is considered low, what is the normal method of getting high resolution tracking? I assume your suggesting direct (1:1) coupling of the encoder to the axis.

If I could think of a use for it my encoder could resolve to 1 part in 16,777,216 +/- one digit

Ian P

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Ian Phillips on 04/04/2014 21:45:07:

I don't know much about telescopes and astronomy but if a 4096 increment encoder is considered low, what is the normal method of getting high resolution tracking? I assume your suggesting direct (1:1) coupling of the encoder to the axis.

.

 

Ian,

Decent Telescope mounts typically use Worm drives … and the bigger the ratio the better.

MichaelG.

.

Edit: Here is a rather nice example of a home-made Equatorial 

 

Edited By Michael Gilligan on 04/04/2014 23:08:36

[Neil WyattModerator @neilwyatt]

Hi Ian,

4096 steps would be good enough to point a telescope in the 'right direction', but 1/4096 of a day is about 20 seconds. Even through my camera I can see things moving across the sky and the moon moves about 1/6 of its disk in 20 seconds so it would not be good enough for tracking.

Michael,

I've been playing with the idea of buying an 8" parabolic reflector and making my own mount (put the cash into the optics) but I hadn't thought of making something quite as hefty as that!

Neil

[Ian PParticipant @ianp]

Neil

Directly coupling the encoder to the axis would give the results you showed but I would not think that is how its normally done. Unless you are on top of a mountain you wont use anywhere near even half of one revolution of the encoder. Presumably encoders when used would be on the shaft of a highly geared drive motor, Correctly geared the encoder I have would resolve to 2-1/2 hundredths of a second.

Photographs would still be blurred as I assume long exposures are the norm.

If you are taking pictures whilst a telescope is tracking under power is the speed constant (once calibrated) or is it servo controlled in some way from the object being imaged?

Ian P

Edited By Ian Phillips on 05/04/2014 20:52:09

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Neil Wyatt on 05/04/2014 15:13:13:

… making my own mount … but I hadn't thought of making something quite as hefty as that!

.

Rather lovely though, isn't it

Meanwhile … you wll probably find this interesting.

MichaelG.

[Neil WyattModerator @neilwyatt]

HI Ian,

most scope mounts use steppers so an encoder is redundant. My thought was an encoder direct coupled to the scope would allow you to manually re-target and let the software know what you have done. I don't know how fast you can operate that unit but it could be possible to gear it up.

Neil

[Johnboy25Participant @johnboy25]

Ian…

what a nice bit of kit this is! You could always make a DRO with it. Dividing the output down to a more usefull pulse per rev number using an Arduino micro controller board and a bit of code….

John

Edited By Johnboy25 on 06/04/2014 20:46:11

[Ian PParticipant @ianp]

Posted by Neil Wyatt on 06/04/2014 18:39:41:

HI Ian,

most scope mounts use steppers so an encoder is redundant. My thought was an encoder direct coupled to the scope would allow you to manually re-target and let the software know what you have done. I don't know how fast you can operate that unit but it could be possible to gear it up.

Neil

Max encoder RPM is 6000, but full electronic operational performance is limited to 3000RPM.

Purely out of interest, what sort of reduction ratio is typical between the stepper and the axis?

Ian P

[Neil WyattModerator @neilwyatt]

THE IDEAL

If I make an EQ mount for my camera, ideally I need it to move with a resolution of better than 1 pixel. Even with the reduced resolution of the display I can still watch objects move across the background in real time.

The moon comes out about 1200 pixels across a diameter. 1200pix ~= 0.5 degree.

360 x 2 x 1200 = 864,000

So ideally around 2 million steps per revolution. I have a big stepper that can do do 1.5 degrees or 240 steps, so that would need to be geared down by 4000:1

That would require the stepper to be driven at about 23Hz.

IN PRACTICE

In practice viewing is never perfect, especially using a camera, and even stars show up as a small group of pixels, not single pixels. Exposures of over a second and stars start to appear as short trails. Perhaps 1000:1 gearing and a ~5Hz drive rate would be adequate.

All of which sounds do-able, but it might be expecting a lot of the quality of the drive train to keep it smooth!

So I resort to google and find:

**LINK**

This give 20-30 arc-seconds as typical which is 64,800 steps, but claims to give a result of 2.5 arc-seconds using feedback from an optical encoder, which is roughly where me 'ideal' would be sitting.

Can anyone elucidate further?

Neil

[Andy AshParticipant @andyash24902]

From that, I would recommend making an electronic indexing rotary table.

I guess you would need a stepper motor too.

I'm not a huge fan of CNC. I know it has uses and does things I couldn't on a manual machine.

For setting out holes accurately on a PCD, machining flats, or gears.

An electronic indexer must be king.

Touch on, type in the number, press the index button, drill the hole, cut the tooth, whatever!

[Ian PParticipant @ianp]

Andy

If I used a stepper I would not need the encoder (assuming its in a well designed system that does not miss steps).

What I was trying to find out with my original question, is what piece of workshop equipment would make full use of the this encoders ability to know its absolute position (to 1 place in about 17 million) even after its been switched off.

I think all of the suggestions so far could be done with an incremental encoder.

Ian P

[Author]

Posts