[Biggermens]

Hello there

I am trying to figure out what determines the resolution of a Glass linear scales ?
Is it the glass, the read head or both ???
Is there anyway to tell by eye ??

Thanks

[Jim Dawson]

Both. The read head has electronics the interoplates the pulses from the lines on the glass. There is no way to tell by eye. The easiest way is to connect the scale to your machine encoder input and see how many pulses it generates in 1 inch or 25mm of movement.

[Al_The_Man]

It is quite a unique process, the typical linear scale or rotary incremental encoder, which uses a similar method by using a read head that uses photo electric sensor, the moving head carries a small section of glass scale with the same resolution as the main scale, but with a slight tilted angle, this creates something called the Moiré effect, which essentially creates a shutter which essentially magnifies the minute resolution/increment of the scale in order for the read head to register the change in pattern.
Otherwise the typical resolution of a scale such as say, a 2µm scale, would not be able to be read by a normal photo sensor.
So the bottom line, as Jim said, it would be impossible to tell by eye.!
Al.

[rallyx99]

It is quite a unique process, the typical linear scale or rotary incremental encoder, which uses a similar method by using a read head that uses photo electric sensor, the moving head carries a small section of glass scale with the same resolution as the main scale, but with a slight tilted angle, this creates something called the Moiré effect, which essentially creates a shutter which essentially magnifies the minute resolution/increment of the scale in order for the read head to register the change in pattern.
Otherwise the typical resolution of a scale such as say, a 2µm scale, would not be able to be read by a normal photo sensor.
So the bottom line, as Jim said, it would be impossible to tell by eye.!
Al.

Is this the same with magnetic scales? ...Are there little tick marks that get picked up magnetically somehow?

[Jim Dawson]

Is this the same with magnetic scales? ...Are there little tick marks that get picked up magnetically somehow?

Yes, that is exactly how they work. The magnetic lines on the magnetic strip are normally spaced at 2mm, the two sensors in the read head somehow sort it all out, and output a pulse train with resolutions as small as 1 micron (or smaller on the high end units).

[Al_The_Man]

In the optical version. the 'trick' is the Moiré effect, the lines that are spaced down to 2µm is impossible for the typical photo-cel to distinguish between such a fine grating, the small piece of grating that is carried by the head is tilted slightly to create a much larger shutter effect.
In some cases you can see this under strong light in a rotory encoder by turning it slowly, the 'shutter' created is about the same width as the grating height, thus enabling what would be an impossible task for a read head to distinguish such a fine resolution.
Incidentally, the 'shutter' rotates at right angles to the direction of head motion.
It is quite neat to see in person.
The quadrature, 90 deg, shifted pulses can be read on all four edges, thus magnifying the quadrature pulses x 4.
In the magnetic head, often a different system is used using the co-tangent difference between the two quadrature sine wave pulses.
Al.

[rallyx99]

Yes, that is exactly how they work. The magnetic lines on the magnetic strip are normally spaced at 2mm, the two sensors in the read head somehow sort it all out, and output a pulse train with resolutions as small as 1 micron (or smaller on the high end units).

That’s pretty interesting! So based on that, it seems like it should be possible to make the read device miss lines by moving it really fast in relation to the scales. Is this correct?


Sent from my iPhone using Tapatalk

[Jim Dawson]

That’s pretty interesting! So based on that, it seems like it should be possible to make the read device miss lines by moving it really fast in relation to the scales. Is this correct?


Sent from my iPhone using Tapatalk

Depends on what you mean by ''really fast''. A 1 micron resolution scale will read reliably at about 4M/sec, with the standard read head, and as fast as 37M/sec with an optional read head. 4M/sec (~17 ft/sec, ~12,000 IPM) is much faster than I would ever run a normal machine tool.

[Al_The_Man]

Probabally comes down to the frequency response of the read head components, the rate of response of the controller, and the form of transmission.
Al.