[playinsteel]

While driving down the road a couple of days ago, I had an idea for a cheap DRO setup. Now this isn't going to be a super accurate system, but I need to get something setup to do some basic mill work.

While surfing the net a while back, I came across a site that used a magnet, a reed switch, and a calculator to give very basic automated counting for a machine used to wind guitar pickups. I thought of modifying this to give a positional readout for an x-y table.

Materials:

12 - small magnets. I found these at Radio Shack for about $2.00 for a package of 6.

2 - SPST Reed Switch. Again, a Radio Shack special, around $3.00 each.

1 - Adhesive of some type. I am using construction cement because I had some lying around the shop.

2 - Cheap calculator from your local discount store. I found some fo $2.75 each.


Construction:

OK, I am starting from scratch. I am building my cheap version of a mini mill that incorporate the DRO design. The x-y table is being built from MDF, 1/4-20 threaded rod, and some homemade antibacklash nuts. Most of this design has been "stolen" from the cnc forums.

After the table is made, the next step is to make a disk that can be mounted on the allthread at a point that will not interfere with the movement of the table. I haven't decided on final deminsions, but probably about 2" in diameter. This disk will serve as the mount point for the magnets, 6 on each axis spaced at 60 degrees apart.

Now the mounting of the reed switches. These must be fixed at a point that is perpendicular to and in close proximity to the magnets on the disk. This should be fixed so that each rotation of the threaded rod will generate 6 cycles of contact from the reed switch.

Next, we'll attach wires to each lead of the reed switch. The other end will be run to the calculator. Open up the calculator and locate the "=" pad on the pcb. Solder the other end of the wires to the "=" pad on the calculator, one to each side of the pad. Now each time a magnet passes the reed switch, it will close, completing the circuit and acting like someone pressing the enter key on calculator.


Usage:

Now we have to figure out the linear movement of the table per revolution of the threaded rod. I am using 1/4-20. So each revolution provides 1/20 inch of movement or approx. 0.05 ". But remember, we have six magnets on the disk, so we will be sending 6 pulses for each revolution. That makes the math go something like 1/(20*6) or approx. 0.0083".

So this gives a theoretical resolution of about eight one thousanths of an inch. Now when we account for errors in my ability to accurately build the table, inherent backlash in the threaded rod (even with the a/b nuts), and other factors, the real world resolution will be somewhat less.

To actually use this, you need to setup each calculator so that it knows how to increment the count for each pulse received from the reed switch. So, simply turn them on and enter the following: 0+0.0083= and you should now be showing 0.0083 on the display. Now, as you turn the handwheels on the mill, you should see the display start to increment as the axis moves. This should give a DRO that is accurate to somewhere between 1/32 and 1/64 inch. I haven't finish this, so I won't know for another day or so how well this works, but maybe this will give rise to a discussion on how to further refine the homemade DRO and inexpensive alternatives to fairly accurate positioning on a budget.

Drawbacks:

This design is limited in accuracy, but should be better than freehand milling.

The initial setup of the calculators causes the count to be off by 0.0083", so you may want to compensate for this by adding an extra 0.0083" (or one additional pulse) to the final measurement. However, I don't think with the limited accuracy this will be a big issue.

The calculators do have a limit to how fast they can process the pulses from the table, so motorizing the axis might not be possible, unless you use gear reduction to lower the speed of rotation of the threaded rod. Again, I haven't been able to test the actual limits on this, but I will report the findings as soon as it's complete.

I'll try to give more detail, dimensions, etc. as I build and finalize the design.

All comments, opinions, suggestions are welcome and encouraged. Thanks for taking the time to read my crazy idea.

James

[pen25]

im thinking if you ise an old roller mouse.. remove the encoders and use them. think that would work?? one other thing would be to create one of those ultrasonic tapes for proximty sensors.. something to think about..

[playinsteel]

im thinking if you ise an old roller mouse.. remove the encoders and use them. think that would work??

You know, I thought of that, but wasn't sure what it would take (in terms of other electronic components) to make it work. In fact, I have a couple of old serial "mice' that are just lying around in my junk pile. I will probably look at doing this, after i finish this project. My goal with the calculator DRO is a cheap (< $20) unit that actually works and is accurate enough for most basic milling needs. Another drawback I just thought of.... this design will not automatically detect if you change direction, so the calculator will have to be set to subtract 0.0083 when direction is changed. I know this is a pain, but I guess theres a tradeoff for cheap and easy.

Thanks for the reply, and I'll definately be looking into the mouse encoders in the near future.

James

[hoagiem]

check out the dro at micro mark. it is the same concept you are talking about.