The Ziman split HP/Agilent interferometry lasers are easy and cheap-enough to find on eBay, but the control boxes are rare, expensive and bulky.
So Sam Goldwasser and I made a replacement that connects to the PC via USB or to your phone via bluetooth.
It does displacement, velocity, angle, straightness measurements and a real-time frequency decomposition.
[CNCzone interface hangs every time I try to add an image here...]
Development history here:
https://sites.google.com/site/janbeck/home
Sam documented the thing here:
Micro Measurement Display 1 (µMD1)
and he also forked the firmware to add sub-wavelength interpolation to single-digit nanometers (you lose bluetooth support there. Alternatively, the Windows software can do interpolation via oversampling).
I think he was thinking of selling kits, but I don't know what became of that.
There are a few people around the world that have built this so far, but my purpose for making it was CNC machine calibration. So I hope it's useful for the crowd here.
Btw. the hardware cost in addition to the laser is <$50 !
Similar Threads:
Wow, we must have 20 of those old HPs and suitcases full of optics at work. No one wants to use them any longer since the renishaw software is so much easier. Great project!
Sent from my A3-A20FHD using Tapatalk
We have the MP10 renishaw, and the new one. The hp's have the little readout box, and some arcane software that nothing can run anymore. The renishaw software is just so easy to use, set up start and end points and number of targets, and decimal accuracy if you want, then start your program. It gives a nice graph and a positional deviation chart. The entire set up is easier with the weather station and such. We are going to buy the Xd from api laser this year, that does position, squareness and flatness with one set up, that takes half or less time to set up than even the renishaw.
Sent from my A3-A20FHD using Tapatalk
Here is a bad YouTube of it
Sent from my A3-A20FHD using Tapatalk
Xd with all the options will run about 50k. We run the g-code program on the machine, it is just a position program with a dwell at each interval, the renishaw is set up for a slightly less interval, so let's say 5 second dwell and renishaw set at 4. You can tell the renishaw software a measurement range and it will average for that 4 seconds then plot the data. Dwell and position depend on what you want on the end result. Some machines we are only looking for .0001, so that can just be a second or two. Other machines we measure down to .000001, that takes a longer and more patient run, fortunately those are small machines.
Sent from my A3-A20FHD using Tapatalk
It is interesting to me how this is used in practice, so I hope you don't mind a few more questions. This system will give you roll, pitch, yaw and distance (and squareness if you put the pentaprism in) along those dwell points in a line, right ? So do you have to move mirrors or anything to acquire a second line?
Since I don't have one yet, can't really answer 100%, but from my limited knowledge the laser itself shoots 2 beams and one gets split. The receiver itself doesn't refract the beam, it wirelessly transmits the data back. Big seller for me on this system was a demo they did for us. We had set up the renishaw on a very long bed lathe, 20ft bed. Took us 8 hours to get the renishaw squared up and aligned using the HP optics. They came in, and by the time the laser pre heated, it was aligned. The only downside is its not quite as accurate (decimal place) as the renishaw, but 90% of what we do is over .00001 anyways.
Sent from my A3-A20FHD using Tapatalk
Very interesting - in their 6D sensor patent, they are using a single frequency laser, with a clever technique for roll detection.
In general, how do you fit the information the laser gives you into the machine tool? Are you actually using roll, pitch and yaw, or just x,y or z distances to calibrate that lathe?
Using it to manually comp the lathe, pitch error and straitness comps in the control, however, we prefer to use the mechanical route to square things back up if possible, and that can include grinding scraping or lapping. If pitch error is too bad, we send the ballscrew (the machines that use them and not linear motors) to a rebuilder, as we can't grind screws on site.
Sent from my A3-A20FHD using Tapatalk