What's new
What's new

DRO system WITHOUT A SCALE,is it possible?

edwin dirnbeck

Hot Rolled
Joined
Oct 24, 2013
Location
st,louis mo
The troublesome part of INSTALING and MAINTAINING a digital readout system is the SCALE.The read head is usually small and simple.Could a read head be replaced with a high resolution electonic camera? The camera could be focused on a section of the machine that would have a random scratch pattern.Wherever the machine would stop ,The camera would see an image that corresponds to ONLY THAT POsITION.Would this be practical or possible? Edwin Dirnbeck
 
Until it sees a chip or other anomaly and reports back the distance to Jupiter. If you seal it well enough to keep that from happening then you have linear encoder which you seem to be avouding.

Ed.
 
I dont think so. Biggest problem I think, if you are using random features already present then how would you calibrate their location? (as accurately as a scale would)

Magnascale makes a neat magnetic linear encoder. The reader head is small and has a hole through it which a small wire is passed that has alrernating magnetic areas(looks like a 3/32 TIG rod)And I do not think the wire is magnetic, idk how it works. Point is, thats what we used in harsh environments, even underwater.
 
Yeah it might be possible, but calibrating it would be a bitch. It would have to learn the existing surface texture and would probably require some pretty hefty computing power compared to what the systems need now. In other words it would be a step in the wrong direction I think.
 
I dont think so. Biggest problem I think, if you are using random features already present then how would you calibrate their location? (as accurately as a scale would)

Magnascale makes a neat magnetic linear encoder. The reader head is small and has a hole through it which a small wire is passed that has alrernating magnetic areas(looks like a 3/32 TIG rod)And I do not think the wire is magnetic, idk how it works. Point is, thats what we used in harsh environments, even underwater.

I think that if you move the camera slowly to the left there will be a situation where MOST,(90%?) of the pixels will See the same image as the pixel to its right had before moving . At this point,you have moved your camera the distance of one pixel spacing.This is a very accurate and known distance. This image could be put in an image library and would be associated with an axis position .This process could continue for the full length of travel as an initial calibration. Edwin Dirnbeck
 
I think that if you move the camera slowly to the left there will be a situation where MOST,(90%?) of the pixels will See the same image as the pixel to its right had before moving . At this point,you have moved your camera the distance of one pixel spacing.This is a very accurate and known distance. This image could be put in an image library and would be associated with an axis position .This process could continue for the full length of travel as an initial calibration. Edwin Dirnbeck


The surface it will be looking at needs to be protected from swarf wear and dust So you end up with a box with a small camara inside So a scale like thing anyhow

Peter
 
Do not confuse resolution with accuracy. They are not the same. The issue has been and is indirect position data. In an ideal world position feedback should come directly from the part being machined, but that is not possible. So the feedback always comes through indirect means. It does not matter whether it comes from a glass scale, magnetized rod or a cable or band driving a rotary encoder. All these are still indirect. Now there are band-aids that are and have been applied, like temperature sensors strategically positioned and history algorithms to compensate, but these are nothing more than calibrated best guesses. Then there is economics. At the moment, there is nothing better than glass scales for the buck.
 
The troublesome part of INSTALING and MAINTAINING a digital readout system is the SCALE.The read head is usually small and simple.Could a read head be replaced with a high resolution electonic camera? The camera could be focused on a section of the machine that would have a random scratch pattern.Wherever the machine would stop ,The camera would see an image that corresponds to ONLY THAT POsITION.Would this be practical or possible? Edwin Dirnbeck


WHY ON EARTH WOULD YOU WANT THAT?

First an optical reader is just that.. (glass scales have a camera that reads the lines) so why would you want a high res camera just ridiculous. think oil on the lens, or coolant.

Next whats the big deal with a dro scale? KISS is great...
 
There are sensors out there (ask MotionGuru), I searched a few, can't recall the
terminology.
Using a laser or other beam to go out, bounce off, and come back, provide
distance to target. I think vibrating mirrors were involved.

The one I looked at maybe 10 years ago was 4' max. distance, and .004" resolution.
 
I suppose a maker could laser engrave a scale on the cast iron at the factory. I think they make magnetic tape that a reader head can use. Not sure if it is just held on by magnetic force or what.
Bil lD.
 
I suppose a maker could laser engrave a scale on the cast iron at the factory. I think they make magnetic tape that a reader head can use. Not sure if it is just held on by magnetic force or what.
Bil lD.

Bill, that would be still indirect feedback. The table is not the part. Consider for a moment that these glass scales are made the same way circuit substrates are etched in the chip making business. Their accuracy is measured in microns. Great headway has been made by IBM in vertical magnetic domain recording for even greater data density per linear length, but that technology has not made it yet to DRO's and even if it did, it would not be for same price of glass scales.
 
Edwin, you are describing how optical mice work. Many mice claim >10000 dpi. This one claims 12,000 for under $30:

Corsair HARPOON RGB PRO FPS / MOBA Gaming Mouse, Black, Backlit RGB LED, 12000 dpi, Optical - Newegg.com

The actual sensor is much smaller that the mouse. Perhaps you could start there.

Wheels 17, BUT BUT BUT so expensive . to complicated .,what if it sees a speck.ect.A common iPhone has 20 million pixels,so if a chip blocks 4 or 5 million pixels,I think a computer could figure it out.Thank you for the THOUTFUL and interesting information.I have laptop and don't use a mouse,so I was not familiar with them.Thanks again Edwin Dirnbeck
 
Installing a good DRO with scale is a lot less effort than what you're expending on this mental exercise! As for maintenance, mine have gone many years with an occasional external wipe down. Just DO it!
 
I suppose a maker could laser engrave a scale on the cast iron at the factory. I think they make magnetic tape that a reader head can use. Not sure if it is just held on by magnetic force or what.
Bil lD.

YOU DONT NEED A SCALE. The first optical mouse needed GRID PAPER TO WORK. The newest ones use a CAMERA YES A CAMERA to READ ALMOST ANY SURFACE.They use there own pixel spacing to tell where they are. Edwin Dirnbeck
 
With only a few years (or many decades) of cameras and machine vision use I am so confused here.
Heck yes on wire wire bonder to weld that lead I could put you within a few microns with a tube vidicon camera back in the early 80's but that not the same as a machine tool.
Bob
 
look up LIGO (gravitational wave detector). using laser interferometer able to detect something like 1 in 34 trillion (ok just guessing but its that crazy). no scales.
(did cost over one billion dollars tho...)
 








 
Back
Top