One, Two, or Three Axis Laser Interferometer Displacement Measurement System
Installation and Operation Manual

Version 1.26 (26-Nov-22)

Copyright © 1994-2022
Samuel M. Goldwasser
--- All Rights Reserved ---

For contact info, please see the
Sci.Electronics.Repair FAQ Email Links Page.

Table of Contents


This manual provides installation and operation information for the interferometer laser displacement measurement systems parts provided in the Hewlett Packard (HP) Laser Measurement Systems - Hobbyist Specials.

This guide deals specifically with setting up the laser, interferometer, optical receiver, and their power supplies. Complete information on wiring the chipKit DP32 or SG-µMD1 board to the REF MEAS signals and running the µMD1 Windows GUI may be found in the µMD1 Installation and Operation Manual. (Note: This and the links below open in a single new tab or window.)

Principles of operation

More information than you probably need or want to know on this technology may be found at:

The following is just a summary.

The technique here is based on what's known as "heterodyne interferometry" which utilizes a two-frequency HeNe laser and special optics to precisely measure changes in distance (called "displacement") down to the nanometer scale. Other types of measurements including those for angle and straightness can be performed with appropriate interferometer optics.

In a nutshell, the two-frequency laser sends out a pair of superimposed beams that differ in optical frequency by a relatively constant amount (called REF), one polarized horizontally and the other polarized vertically. REF is typically in the low MHz range which provides a convenient "carrier" for signal processing. Optics separate the two beams, sending one to a (generally) fixed reflector, and the other to some tool or device whose position needs to be measured precisely. Both beams then return to an optical receiver where another difference frequency (called MEAS) is generated. With no movement, REF and MEAS have the same frequency and the phase relationship between the two beams is constant. But if the tool or whatever moves, the frequency and thus phase relationship between REF and MEAS will change due to doppler shift. With the simplest interferometer optics, a phase change of 360 degrees represents a position change of 1/2 wavelength of the light from the laser, approximately 632.8 nm for the HeNe lasers most often used. The precise wavelength for these lasers is specified to 6 decimal places (e.g., 632.991372 nm for the HP/Agilent 5517B) with a wavelength accuracy of 0.1 ppm over the life of the laser.

Using the two-frequency approach rather than a basic Michelson or similar interferometer, among other things makes these systems more immune to misalignment by eliminating issues of fringe counting and direction, and fringe contrast, and they are less susceptible to changes in signal level as the laser ages, dust settles on the optics, or alignment changes.

The diagrams below show the organization of a single axis system using µMD1 or µMD2 for the processing and measurement display. Both provide capabilities similar to those of an HP-5508A Measurement Display - and a lot more. For 2 or 3 axes, the interferometer optics, remote (Test Arm) reflector, and optical receiver would be replicated, along with 1 or 2 non-polarizing beam-splitters to divide the output of the laser.

By default these systems will come with µMD2. However, while the µMD1 kit is no longer available, the SG-µMD1 PCB and Digikey "Cart" with all the requireed electronic components may be substituted at slightly lower cost..

IMPORTANT: The laser included in this kit is currently a 5517, but info is included here for the 5501B as well. The differences are as follows:

The photo below shows the typical parts included in a single axis system using a 5517 laser:

For the 5501B version, the connectors and power packs will be different but all other parts are essentially the same.

Styles of some of these parts may vary.


Photo of Typical Parts for Single Axis System using 5517 Laser (Left) and 5501B Laser (Right) with Original µMD1
The interferometers shown are the 10702/6A PBS with 10703A CC, 10722A QWP, and generic CC

For a 2-axis system, a non-polarizing beam-splitter is necessary to divide the output of the laser into two approximately equal parts, and a second interferometer and optical receiver are required. For a 3-axis system, an additional set of similar parts is required. The beam-splitters may be 1:1 or 2:1. Either is acceptable. The power won't be equal but the laser should have enough power. The µMD1 and µMD2 boards can handle up to 3 axes.

Note that the laser included with the single axis system may be lower power than with the 2 (or 3) axis system, but should still have enough power for these if desired.

Wiring power and signals to the laser and optical receiver

The wiring can be done first, or after the interferometer optics are mounted. However, even coarse alignment will be a lot simpler if the laser and optical receiver are powered.

Power supplies/wall adapters

CAUTION: ONLY change connections (including plugging or unplugging stuff) with power OFF! Otherwise bad things may happen.

The laser requires +/-15 VDC. Therefore, a pair of power supplies are provided.

There may also be an option to substitute a cased +/-15 VDC switchmode power supply for the power packs. Wiring of that should be self explanatory as all the connections are labeled. The only "gotcha" is that the two outputs may be totally isolated so that the common point will need to be tied together. Take care when wiring this supply, especially the line voltage inputs. Add strain reliefs or tape so that wires can't be ripped out.

Unless you opted for the optional HP cable, it will be necessary to wire up the laser head connector to the DC power packs and REF input to REF header on the µMD1 board. With the optional cable, two sets of wires from the cable are built-in and they just need to be attached to the power supplies and REF/MEAS on the µMD1 board.

One of the circular connectors provided with kit will fit the POWER connector. The other will probably need to be modified to fit the REF connector as the standard connector must have its inner shell rotated by 45 degrees to fit. If the two connectors differ, the one for REF will have pins that are not installed, but they may both be like that. The inner rubber shell that holds the pins needs to be pressed out and rotated 45 degrees. I usually do that by trimming around the edges with a thin blade to free it up and then carefully pushing it out using a rod or the butt-end of a large drill bit on a drill press. Then rotate the shell by 45 degrees in the correct direction and push it back in. After confirming it's in the proper orientation so the labeling is correct, some rubber adhesive can be added to secure it. But that's usually not needed as it will be reasonably tight without it. Alternatively, or until you get up the courage to abuse the connector, inserting pins for REF and REF- will work fine. The GND is not even needed as long as there is a GND running to the chipKit or SG-µMD1 board from the PWR connector(junction of the - on the +15 VDC power pack and + on the -15 VDC power pack).

10780A/B/C optical receiver

The mating connector for these is a strange 4 pin BNC, which is unobtanium for less than 3 arms and 2 legs (if one can find it at all). Therefore, where a mating connector has not been provided, two male and two female pins/connectors/headers have been included that will mate with it. The 2 wire power cable will probably already be plugged into the 10780 and the color of the wires is correct - red for +15 VDC and black for GND/common. Put heatshrink over them once attached to wires to prevent shorts and secure with hot-melt glue if desired. They should not need to be removed.

   BNC Pin   Function
  1 (LL,F)    ~MEAS (Zeeman beat signal pair from
  2 (UL,F)     MEAS  differential line driver.)
  3 (LR,M)    Return (also BNC shell and receiver case.)
  4 (UR,M)    +15 VDC

       |     |
       |     |
       | TP  |
       |     |
       |     |
  MEAS | x o | +15 VDC
 ~MEAS | x o | Return

(Sometimes the connector insert gets rotated slightly. But the "x" denotes a socket while the "o" denotes a pin.)

If there are a pair of wires sticking out of the 10780, red is +15 VDC and black is GND/return. The pins are for MEAS/~MEAS.

If you are fortunate enough to have received a mating connector, or better yet, one with a piece of cable attached, use a DMM to confirm the connections since the color coding varies.

To avoid ground loops, the case of the 10780 should NOT be connected to anything - use insulating "hardware" to secure it to the optical table or whatever. The 10780 originally comes with rectangular black plastic insulators, but these often have a habit of disappearing after awhile. Nylon washers with Nylon screws work just as well.

There may be a gain trim-pot accessible on the top of the 10780 (B version and higher). If response is erratic, it can be adjusted. Usually they can be turned all the way up. If that results in instability, just back off until it quiets down. If the optical receiver is a 10780A, the gain adjustment is internal. Normally, these do not need adjustment. Regardless, it will probably be fine without adjustment but if the laser locks (READY on solid) but there is no output (or signal LED) on the 10780, the gain may be set too low. Check the REF signal from the laser either with a scope or the REF readout on µMD1 to confirm that it is actually locking correctly. Around 100-120 µW should be sufficient for the laser to lock; the 10780 requires under 10 µW to produce a valid signal. Also, if you have a 10780F or 10780U instead of a 10780A/B/C, it has no lens and no polarizer so there would be no signal unless the laser power were higher and a linear polarizer at 45 degrees was added in front of the optical window.

"TP" is the single pin via a feed-through used to monitor signal level during interferometer alignment. Its output is non-linear, being compressed at the upper end with strong inputs. Monitoring it with a DMM or scope is useful during interferometer alignment.

Initial testing

Double check your wiring and then apply AC power to both power supplies and measure their output voltages with no load. Their outputs should be close to +/-15 VDC. (The positive supply used with a 5517 may be slightly higher, but once a load is connected, it will drop to an acceptable value.)

Remove AC power, wait for the voltages on the supplies to decay, and then connect the laser. Turn power back on. Immediately, 3 of the 4 LEDs on the back of the laser should come on - all except READY. Anywhere from instantly to awhile after that, a beam should appear. Make sure the turret at the front of the laser has its large hole at the top.

IMPORTANT: Some lasers may require up to a minute or possibly even more for the beam to appear. While the laser in the kit is probably high mileage, even youthful $10,000 lasers may take some time to start. But if there isn't a beam within a couple minutes, power down and double check the power the power supplies and connections. Others may not restart immediately if powered off and back on again or may flicker - wait a few minutes before repowering if this is the case. Failure to do so may result in damage to both the laser tube and HeNe laser power supply brick inside the laser. Non of this helps, contact me.

CAUTION: DO NOT plug or unplug connectors until the voltage has decayed to close to 0 V (indicated by ALL LEDs on the back of the laser head going out). If pins make contact in the wrong order, an internal fuse in the laser may blow at the very least. But it could be worse. It wouldn't hurt to wire an LED+current limiting resistor across each power supply output to monitor this (though the LEDs on the laser will serve when connected).

After a couple minutes, the READY LED should start flashing. With a beam present, after another 2 minutes or so, it should stay on solid. The laser is ready for use. (A minute or more possible for the 5501B.)

Aim the 10780 directly into the laser. (It must be roughly aligned with the X or Y axis to get a signal.) The green LED at the top of the optical receiver should light indicating that it is detecting the f1/f2 difference frequency. An oscilloscope on MEAS or ~MEAS would then show a squarewave at the REF frequency. There is a gain adjustment inside the 10780 which affects the sensitivity, especially for small signals. For the 10780A, it is hidden by the top cover. Usually, there should be no need to touch it, but on the off chance that it had been set too low, loosen the small screws at either end of the 10780A and remove the top cover. It's the trimpot facing up. Turn it fully clockwise. (DO NOT touch any other trim-pot.) At the same time, drill a hole in the cover at the trim-pot location for the future. :) (For the 10780B/C, there is an access hole in the top cover.) Sometimes, there will be instability if turned too high. If a signal is detected with no beam into the optical receiver, back it off.

Connecting µMD1 or µMD2

µMD1 usually comes with headers and mating connectors while µMD2 usually comes with screw terminal blocks, which are handier if cables don't need to be changed. One type can be substituted for the other - parts are readily available.

Cables will need to be constructed for REF from the laser and MEAS from the optical receiver. For runs of a few feet or less, twisted pairs for REF/~REF and MEAS/~MEAS will suffice. The common/grounds should be run to the common point of the power supplies but they don't need to be in the cables and the cables don't need to be shielded. The 150 ohm terminating resistors on the µMD1 board will provide the needed connection so that the inputs to the line receiver are at the proper level.

Complete information on wiring the SG-µMD1 and SG-µMD2 boards and connecting them to the REF and MEAS signals, and running the µMD1 GUI may be found in the µMD1 Installation and Operation Manual and µMD2 Installation and Operation Manual, respectively.

Once it hooked up and live, the built-in frequency counter readouts of the µMD GUI can be used to monitor REF and MEAS. However, an oscilloscope will still be better at detecting marginal signal quality due to poor alignment or other problems. It's best to connect MEAS to channel 1 and use that for the trigger; REF to channel 2. Dig out an old analog scope and put it to good use! ;-) Low cost digital scopes are available that would be suitable - under $100 for one the plugs into USB on a PC.

Wiring additional axes

The same laser will drive multiple axes. Power for the optical receivers for MEAS2 and/or MEAS3 can be taken from the +15 VDC power supply. There is plenty of extra capacity.

MEAS2/~MEAS2 and/or MEAS3/~MEAS3 should be run from their optical receivers to the respective inputs on the µMD1 or µMD2 board. Their grounds do not need to be connected (as long as there is a ground path for axis 1).

Setting up the interferometer

Any of the common interferometer configurations can be used with µMD1 or µMD2. Those for displacement (relative distance) are shown below:

Depending on version, the Hobbyist Special kits include either of the following for each axis:

  1. 10702A with 10703A, Generic CC: Standard Linear Interferometer (LI) with with unmounted CC triheddral prism.

  2. 10702A with 10703A, 10722A, Generic CC: Plane Mirror Interferometer but with one non-HP CC. Also 3/4x3/4 inch planar mirror on cheezy mount.

  3. 10702A with 10703A, second 10703A, 10722A: Standard Plane Mirror Interferometer. A standard HP mount for the interferometer will be included pending availability. Also 3/4x3/4 inch planar mirror on cheezy mount.

With 2 axes, a non-polarizing beam-splitter (10701A) will also be included to divide the output of the laser, and second 10701A for 3 axes.

By removing and/or rearranging the PMI components, these can also be used as Linear Interferometers with remote retro-reflectors.

Where a bare trihedral prism is included for the second retro-reflector rather than a 10703A, if intending to use it in the PMI, it will need to be attached to the PBS cube. This can be done using glue or tape as long as care is taken not to damage the optical surfaces.

For serious applications, (3) is recommended to minimize the clunkiness factor if nothing else. ;-) But by assuring that the generic CC is in the reference path (F2 in the diagrams), (2) should be acceptable for long distances.

Note that the components of the interferometers can be rearranged for right-angle instead of straight-through operation and/or the locations of the laser and optical receiver can be swapped. The only consequence in measurement is that the sign of the displacement may flip. Any configuration that results in a similar beam paths inside the interferometer will be satisfactory. The only restrictions is that all components must be oriented on a multiple of 90 degrees so that the f1/f2 components always align with X or Y. There are ways around this requirement but that's for the advanced course. ;-) And it's best to orient the 10703As (cube-corners) so that the beam does not intercept an edge of the prism. This will depend on both the configuration and whether the beams are above each-other or next to each-other, and can be done visually.

The more compact Single Beam Interferometer (SBI, 10705A with 10704A RR) can also be used along with a remote 10704A RR or equivalent. The resolution will be the same as that of the LI.

Some options for the "Tool" in a demonstration system above:

Aligning the interferometer

Power the laser and optical receiver and wait for the laser to come READY.

With genuine HP/Agilent optics, it's very easy to get all this working together. Even modest misalignment can be tolerated, though the signal quality may degrade somewhat. Once everything is aligned, make sure it's all locked down. Except for the remote reflector (cube-corner or mirror), nothing else should move! Even with the non-HP retroreflector, as long as it is mounted securely, there should not be stability issues.


What trouble? Coming soon.

  • Back to Hobbyist Special Interferometer Table of Contents.

    -- end V1.26 --