FezCNC

Yes, there is a sync pin, but if you add ramp up/down profiles and the X motor must move 200 steps and the Y motor must move 10 steps then the motion profiles are going to make a mess of things… :slight_smile:

I assume that a laser micrometer has a full scale range of -+1mm? That will make it useless for CNC, but can have other interesting applications. I also assume that it isn’t cheap… :slight_smile:

@ kertnell, you can look at Optical Linear encoder, or rulers, they are designed for what you need, and they are expensive, new anyway… I have looked at adding an encoder to my lathe for a digital readout, but was quoted, if I remember correctly, $700 for a 500mm encoder…
See Linear encoder - Wikipedia

I have not read the datasheet fully but I would assume that the sync would trigger any slave drives to a master drive, the master emits the sync and then all start moving at the same time. In the case of a CNC machine you would have to plan the moves sent ot each drive to that they could be cooridinated. For example break a path into n chucnks and in each chunk you know where each axis needs to move and at what velocity. Each axis will be done executing each chunk at the same time. You may have to control the IC in velcoity mode to keep it from wantign to accel/deccel with each small move though.

I just ordered a couple of Eval boards to play with, nobody has them yet but Mouser shows they should have a few about 7/7/2011.

EDIT: Also, using a shared external clock should provide better syncing I would think.

kertnell? Funny “errol”, real funny :slight_smile:

I was thinking to use the the divergence factor of a laser pointer to measure the the current position of the axis. (See image attached)

Lasers spread out over distance; this amount is knowable. This means that if you are further away, the beam spreads out. If then passed through a small hole, and diffused with a colour filter, then you can measure distance with a light sensor. Since I have 24bit ADC’s I can connect it to the photosensor with almost no system noise (since the sensor will be soldered directly to the board, very close to the IC).

What do you guys think?

@ kurtnelle, sorry I copied your name from Jeff’s post… :). Digikey lists the price at ±$13 per chip…

Generally laser micrometers mix the transmitted laser with the received laser. This gives you destructive and constructive interferance, but at the wavelenth of the laser, ie 808nm. Thus you can’t destinguish between 808nm distance and 1616nm as both will give you the same distance.

You can frequency modulate the laser at 100MHz, then do the same mixing. This will give you a full range of 6 meters. But it isnt easy woking at these frequencies…

Yea, I have no idea how it works. I’m assuming that the color (i.e. frequency) changes based on the mixing of the 2 frequencies of light that you are receiving, and that based on the resultant frequency that would be the point at which the light beams are forming constructive and destructive interference.

Since I don’t have a method for determining the frequency of the light; what about my previous suggestion?

No, the frequency won’t change. Two streams of light that are out of phase will interfere with each other and produce specific patterns. Think of splashing both of your hands in the water 1M apart at the same time but one hand hits 1/2 second before the other. When the waves hit they will interfere with each other, the resultant pattern will be be determined by the difference in their phase (time arriving at a given point.)

@ Jeff So then what does the resulting interference look like/percived as? A color change or a picture change?

[url]http://en.wikipedia.org/wiki/Interference_(wave_propagation)[/url]

Don’t let the colors shown in the one image confuse you. The colors represent the peak/trough of the wave NOT the frequency.

When the surface is multiples of 2 wave lenths from the sender/receiver, the going and returning waves will cancel and you will have 0% light output.

When the surface is a mutliple of 3 wavelengths away, you will have 200% light output.

The brightness of the light will change…

But mixing the incomming and outgoing light requires half mirrors and prisms and fancy optics…

(hmm, some of the optics might be salvaged from a CDROM/CDRW drive’s opticts…)

Also take a look at:

@ Jeff So the beams are fired one side the other, and the image of them interfering at the point of contact where they disperse (like splashing water) is what determines the distance?

EDIT: Yep, expensive optics.

There are other types too that use rotating mirror and optics to create a sheet of light which which broken by an object allows the size to be measured. If the object is rotated you can measure its profile.

[url]http://www.laserlinc.com/laserdemo.html[/url]

4 packs of goodness showed up today :slight_smile:

I’ve obtained the bearings for the axis, and figured out the last mechanical problems. Soon the fun will begin.

For those that are interested. I’ve started attaching the steppers to the framework. Still got alot of work to go.

Nice CNC project kurtnelle… I was wondering, have you looked into some “ACME” threads or similar threads; they can move your cnc faster, with more power, and they wouldn’t over heat like the regular threads your using.
keep up the good work. :wink:

@ Alex,

I have encountred extreme difficulty and expense in obtaining acme lead screws, which is why I’m forgoing them for the time being. This CNC’s purpose is to create a pretty cnc that will be used long term.

@ Errol, So it looks like I’ll be trying out the L6470 after all [url]http://www.st.com/internet/analog/product/248592.jsp[/url]. I purchased the Eval [url]http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATA_BRIEF/DM00025052.pdf[/url] for it ($92.86 bucks btw). It has a simpler interface than the Quad Stepper, and I’ll try to make public the driver for it if I end up using it in the design.

The most exciting feature this chip has is it’s stall detection. That could prove useful. I’ve read on the overflow forums that if a stepper stalls and you keep driving it, you will demagnetize the magnets eventually.

it looks like I’ll be trying out the L6470 after all…

You might also want to keep an eye on what Rob Giseburt (who’s pretty well-known in the home 3D printer world) is up to: tinkerin.gs: Motate128

Why is it that there is always someone else doing the same thing as you at the same time as you.

Why is it that there is always someone else doing the same thing as you at the same time as you.

Because great minds think alike?

I suppose that will have to be the correct answer. :slight_smile: The other answer is just depressing