Do-It-Yourself Leveling with a Laser Pen

(article continued from previous page)

Some arrangement of rulers, blocks, clamps, vises, and shims are usually adequate for holding pieces together in a straight orientation during gluing. However, things get complicated when the pieces are curved or cylindrical with soft, rounded, or beveled edges. There’s nothing square, flat, and definite to key off of as a relative base.

Trying to fit an angled collar perpendicular to the sides of a cylinder.

Trying to fit an angled collar perpendicular to the sides of a cylinder.

Such is the problem I encountered when trying to align an angled ring collar onto the beveled end of a flexible plastic canister body. Although the collar could be positioned adequately by eye, it would be a shame to have it crooked on such a visible location on the robot. And, because the collar is to be permanently attached to the shell with epoxy, it isn’t something easy to correct or do over if a mistake is made.

Magnifying Angular Errors Via Long Distances

A tiny error in angle can make a huge difference over long distances. We'll take advantage of that fact to make an extremely precise relative angle measurement tool.

A laser pointer on a flat plate aims at a tall ceiling to measure parallelism.

A laser pointer on a flat plate aims at a tall ceiling to measure parallelism.

Put simply, a laser is going to sit atop the canister and aim at a spot of the ceiling. After adding the collar, if the laser points to a different spot, the collar is crooked.

Here is the test methodology in greater detail:

  1. A laser is attached to a clear plate (plastic or glass).
  2. Alignment marks (crosshair or whatever) are drawn on the clear plate so that the plate can be taken off and put back on in the same spot.
  3. The plate is placed onto the plane to be considered the baseline. In this case, the end of the canister. I chose a molding mark and recycling symbol on the end of the plastic canister to center the plate crosshair against.
  4. The entire apparatus is placed in a room with a tall ceiling (the taller the better). In this case, the cathedral ceilings in my entrance hallway.
  5. With the laser turned on, the entire apparatus is moved to a location on the floor where the laser dot aligns with a squashed bug or other prominent point on the ceiling.
  6. The canister (but not the plate) is taped to the floor with removable masking tape.
  7. Without rotating the plate or moving the canister, the plate is gently lifted up and the collar is placed over the top of the canister.
  8. The plate is put back on such that it presses against the collar. Tilting the plate also tilts the collar. Adjust the plate until the laser dot again matches the icky point on the ceiling. This means the angle of the collar is identical to the angle of the canister top. Any difference in angle would cause the laser pointer dot to be pointing somewhere else.
  9. This exercise is performed a couple of more times to be confident in the repeatability of the results and to practice the method.
  10. Now that we’ve mastered this, the inner diameter of the collar is coated with epoxy and the alignment method is performed.
  11. The entire apparatus stays in place until the epoxy hardens -- leaving the collar epoxied to the canister body in very precise alignment. The bug on the ceiling, however, remains dead.

There’s a pretty illustration at the end of this page if you can’t quite visualize how this works.

Hacking a Laser Pen into a Precision Laser Level

Fancy laser levels are available at most hardware stores. But, I didn’t want to spend a lot of money. Instead, I picked up a laser pen from the dollar bin at a local drug store.

The laser pen has two buttons: one button for the laser and the other button for a blue LED. I’m not sure why they think a single-LED blue flashlight is worthwhile, but the kids like it.

Initially, I wanted to drill a hole in a wood block and simply stick the pen into it. But, pressing the tactile button takes too much force and is bound to alter the alignment of the precision level. Furthermore, who wants to hold down the button the entire time the setup is being leveled? Perhaps duct tape could be wrapped around the pen to force the button down, but the tape would be messy to add and remove each time the laser level is used.

A laser-pointer module removed from a cheap laser pen.

A laser-pointer module removed from a cheap laser pen.

I decided to disassemble the pen and place in the laser module in a more practical technical instrument. After unscrewing everything and removing the batteries, the printed circuit board pulled out with minimal effort.

A button is no more than a spring-returned switch. So, soldering a battery wire to the opposite contact bypasses the button. Applying power turns on the laser without pressing the button.

A laser module in a plastic block with an external power switch and Energizer battery pack.

A laser module in a plastic block with an external power switch and Energizer battery pack.

I machined a fancy clear plastic block from acrylic to hold the laser module. This isn’t really necessary, but the block glows with a funky aura when the laser is turned on.

The laser pointer power supply wires (soldered to the module earlier) are connected to a switch on a breadboard. Other than a Molex connector for the external power supply, the breadboard doesn’t contain any other circuitry or additional components. It is just a convenient way of holding the switch and connector such that they can be attached to the clear plastic block.

The external power supply is a three-pack of rechargeable NiMH (nickel metal hydride) AA cells. The laser pen came with three alkaline button cells, so the external battery pack supplies an appropriate replacement voltage. (The voltage of your laser pen may differ.) The external pack of rechargeable cells is cheaper to operate and will last much, much longer than the coin cells.

DIY laser level made from a clear plastic plate, laser pen PCB, AA batteries, and an overhead transparency.

DIY laser level made from a clear plastic plate, laser pen PCB, AA batteries, and an overhead transparency.

A few holes are drilled in a flat polycarbonate plate so that flathead screws can hold the battery pack and laser in place.

A thin sheet of clear plastic is taped onto the plate to draw alignment marks for whatever the level is going to be centered against. In the future, fancy crosshairs can be printed onto an overhead transparency to provide more professional look.

If the plate didn’t include alignment marks, it would be impossible to lift the plate and then restore the plate to the same spot. The misalignment would cause the laser dot to appear somewhere else on the ceiling even if the collar is actually parallel to the base plane.

Calculating Laser Level Precision with Trigonometry

How precise is this homemade laser level, really?

Don’t freak out, but here comes a little bit of math. I’ve tried to temper it with a colorful illustration that includes a bug splat.

Calculating angular error by measuring the change in laser spot location at a certain vertical distance.

Calculating angular error by measuring the change in laser spot location at a certain vertical distance.

Let’s say the laser level is aimed at a spot on the ceiling 16 feet from the tip of the laser. That’s 16*12=192 inches. After placing the collar on as best as possible, let’s say the laser dot is no more than 1/3 of an inch away from the original spot.

Trigonometry labels the lengths of the sides of a right triangle a, b, and c. The angles opposite of those sides are labeled in uppercase as A, B, and C.

For the purposes of this test setup, the distance to the ceiling is the length of the adjacent side, which is a. The distance from the original spot to the current mark is b. The amount of error is the angle B.

Length a = 192 inches
Length b = 0.33 inches
Angle B = arctan (b/a)
Angle B = arctan (0.33/192)
Angle B = 0.0984 degrees = about 0.1

In this example, the angle of the collar is only off by a tenth of a degree if the laser dot is within a third of an inch from the original target spot. That’s pretty good!

The farther away the ceiling (assuming the laser dot remains fixed to 0.33 inches), the lower the angular error. Or, even better, the closer the laser dot can be to the original spot, the lower the angular error.

In reality, I figure I probably got the laser dot to within less than 0.2 inches of the original spot. But the cheap laser pointer produces a pretty hairy dot, so it’s hard to tell.


>