DIY Laser Interferometer setup

[Duc]

A few months ago I purchased Laser Interferometer DIY kit off ebay to help tune by Fadal Ballscrews since I was unable to find a cheap check master gage. I would have bought a new ballscrew but everyone has been out of metric 4020 Y-axis ball screws till Oct. The seller Sam was very responsive to questions and fine tuning the kit to my needs. This will be a Interferometer using a Two-frequency laser setup.

Requirements
1. Ability to upgrade the system to multiple axis down the road
2. Enough power to measure a Fadal 8030 for a buddy.
In short order he had a system listed on ebay with a high power laser and 9mm beam size for a decent price of $1,430 with 2 day shipping. I was itching for something to do over the 4 day weekend. Kit uses a HP 5517C that has been verified to work and repaired if needed.
https://www.ebay.com/itm/303500535042?hash=item46aa0aa502:g:9r4AAOSwsTxgQ-Wt


Instruction manuals for kit
https://www.repairfaq.org/sam/manuals/hobspcl1.htm
Bag of parts and one laser.


Not a real difficult system to solder together. You will need a fine tip soldering iron and some patience. I only purchased the board parts for 1 axis but the remaining axis's would be cheap.


The next part caused me about 1.5 days of frustration trying to get the system running. After soldering the board together I accidentally wiped the firmware on the controller board when doing a power on test and had to reloaded the file. Sam had everything posted on his website for the firmware and instructions how to load it. Reloaded the file just fine but I was unable to get the system running after aligning the lasers. I'll give Sam credit for the quick responses about the problem. We exchanged over 35 emails in the last two days trying to figure out what the problem was. Multiple signal checks using a DMM and Oscilloscope resulted in us scratching our heads. In the end we found the firmware on the website was corrupt and he was able to send a working one over. Guess I was the first to blow away the preloaded firmware. He has uploaded the correct one to the website now. Gold Star to me

I've uploaded some pictures of the test setup. The hydraulic table is limiting my ability to get everything level and adjusted perfectly. Now that I have the system running I will be purchasing some additional hardware to make the setup easier. I was able to set the backlash on the middle and positive end of the Y axis but need to adjust hydraulic table before completing full motion.

Initially playing with the setup.


Current setup. I machined a custom fixture plate to mount everything down to.


Screen shot from program.




Items to buy to improve setup

1. Rigid tripod
2. Foam for my pelican case.
3. Lab jack or build a mount for the laser interferometer mirror.

Accuracy of the system should be no issue as the display shows. I only somewhat under the principles behind the science but overall the system will have no issues measuring .00001. Even rolling my chair near the setup causes a change in measurement. I believe the accurany was around 6 nano meters which is about 0.00000023622 inches. I will need to verify the exact number.

Few extra links for learning
Renishaw XL10 page
https://www.renishaw.com/en/interferometry-explained--7854

Sam's link. Way above my head for most things. Massive archive of laser info.
https://www.repairfaq.org/sam/laserfaq.htm#faqtoc

I will add more info tomorrow.

 

[Luke Rickert]

Those are really interesting, I "need" a 2 axis setup, and then I need an application for it I studied physics for my undergraduate degree and have always liked lasers and optics but thought interferometers were just far out of budget.

How difficult/expensive is it to get the mirrors, software etc to do more complicated measurements such as angle and flatness?

Luke

 

[Duc]

Those are really interesting, I "need" a 2 axis setup, and then I need an application for it I studied physics for my undergraduate degree and have always liked lasers and optics but thought interferometers were just far out of budget.

How difficult/expensive is it to get the mirrors, software etc to do more complicated measurements such as angle and flatness?

Luke

Software and a cheap mirror came with the kit to run it has a Planer mirror interferometer which has 4x the resolution than a linear interferometer which I was running mine as. I removed one retroreflector from my setup to use as the movable mirror versus the simple mirror that was included. The simple mirror was a son of bitch to get aligned versus the retroreflector mirror.

I believe the software will run angle and flatness provided you have the mirrors.


Hopefully tomorrow will have more time to write.

 

[Duc]

Now take this with a grain of sand since I have never used a commercial laser interferometer but I image they are difficult to tune in also. I dont even want to imagine a 40 meter long bridge mill.

The kit instruction link. Sam's page is really full of info about a ton of lasers and he even responds to emails quick. Just cant say enough good things about the help I received.
https://www.repairfaq.org/sam/manuals/hobspcl1.htm

Now there is a couple different ways to setup the kit. I tried to use the Plane Mirror Interferometer but switch to the linear interferometer setup for ease of alignment. I will be switching back to the Plane mirror interferometer when I buy another retroreflector.


So this was the first setup I used. Adjusting the mirror was absolute bitch to get right. Even small changes could affect it alot.



Small screws on the back of the mirror. Very sensitive to adjustments.


Now to the issue I had with the simple mirror. If the alignment between the laser orientation and the mirror wasnt perfectly perperdicular to the axis of motion then I would receive a fault over a large movement. I made a quick powerpoint picture that might make more sense. Now the angle could be off a little bit and still cause the return beam to move on the receiver to much.


I even tried using a larger mirror on a gimbal setup that I purchased off ebay. Worked better but still a pain in the ass. Also mounted the gimbal onto a linear stage so I could verify distance measurement which worked well.



Switching to a linear setup let me borrow one of the retroreflectors from the interferometer setup. quote from a website about the benefits. "Retroreflectors reflect incident light back toward the direction of the light source, operating over a wide range of angles of incidence." IE I only had to put the reflector into the beam path to get it to reflect back to the same spot. This is what renishaw and other laser interferometers appear to use.


I mounted the retroreflector inside an adapter that came with the gimbal setup.


I removed the retroreflector and planar mirror from the interferometer to form the linear interferometer setup. Also this is where I got the retroreflector to test with in the picture above.

 

[Duc]

I wish I would have taken better pictures while I running the setup. The next time I setup the laser I will try to get pictures to explain the alignment of the two laser beams back to the receiver and why a oscilloscope makes it easier.

Setting backlash: Much easier alignment of system since the travel distance is only ~.300 inch total at each location. I was able to knock this out quick. Was easy to reset the distance on the computer screen before continuing with the program. Backlash could be read off the screen.

Simulated Example below
Yes mine was this bad. More on that later. Was very easy to verify without having to look at a small indicator inside the mill.


Ballscrew comp: From what I could find online this can be a pain with a check gage or laying down a height gage. Once the laser was setup I could run this in 10 mins. I put some pauses in my program so I could calculate the offset in my head and put it down in a excel file.

Fadal program
%
O8005
N1 G90 G1
y-9.0 F60. (Where I set the computer program to zero, Found a similar example in manual)
G4 P10000 (Delay the next movement so I could record number. 10 second pause)
y-8.0
G4 P10000
Y-7.0
G4 P10000
Y-6.0

etc...

G4 P10000
Y6.0
G4 P10000
Y7.0
G4 P10000
Y8.0
G4 P10000
Y9.0
G4 P10000
Y10.0
N10 M30
%

For a example if I was at -8 the screen would look like this with a following error of .0011

at -8 the screen would look like this with a total following error of .0033. This was a compounding error.


Now most mills wouldnt have this bad of error but my Y-axis ball screw is really bad. Run hard for 20 years and put away wet. Ive been to looking to replace the ballscrew but they are out of stock everywhere. Using this program my final comp numbers came out to the following. Normally a person would use a check gage to find the error at each location then add the value to the prevision error total for each position.

yes it really was out 6.4 thou at full travel. This is with backlash already fixed.

Rough youtube showing the new positional accuracy of my ballscrew. I need to redo this video one day.