Building a X-Y galvo scan setup for laser marking

[macona]

So I have two sets of galvos and drives as posted here: http://www.photonlexicon.com/forums/...t=macona+galvo

I would like to build a setup for doing laser marking with these. I do have a couple f-theta scan lenses, one a short 50mm focal length Photon Gear model for 532nm and a huge one made by melles griot that I am pretty certain is for 1064.

Figuring I use my Lightwave 7 watt average @ 10kpps 532nm head and maybe put a 1064 head together out of some other stuff I have.

What can I do to control a setup like this? It seems like I would need a pretty high resolution DAC to get good resolution of the scan. Can I use something like Scanlab's RTC series boards and software to control galvos that are not made by them?

[krazer]

When I built a similar system with my yag, I used a pair of cambridge 6450's and a matching f-theta lens of out a lumonics lightwriter, and gsi 'lightning' amps.

To drive it I am using the etherdream (only 12 bits, but that is still a million points which even at 10khz takes a while), and a custom program written in c++/gtk which loads in a bitmap image and feeds the points to the scanner. Currently the program is in pretty sorry shape, it can't handle images much bigger than 300x300 points before it runs out of memory and dies and most of the buttons don't do anything, but it does the job


(that is on a soda can, about 1cm tall)

[macona]

That is pretty cool.

How were you driving the q-switch?

[krazer]

I was using one of the color channels (red IIRC), and outputting an 'on' point whenever I wanted a point. I then ran it into a function generator set up to output a 1us pulse at the rising edge of the red signal (experimentally determined to be about the right length for best pulse output), which went into the q-switch driver input. Since your head has a pulse generator built in the function generator shouldn't be necessary if you go that route.

I did not have any pre-pulse kill, so I just ran at a 1kpps so that the crystal was 'fully charged' for each pulse. For a lot of applications it is a bit too much energy (about 10mj), but with the 14" f-theta lens I have a pretty large spot size so the extra power is nice for cutting through soda cans. I ended up using to it to etch pictures into red cups so people didn't get them mixed up, and larser etching logos into some stainless gardening tools for a friend (sadly, don't have pictures on me), and then moved across the country for school so now the rig is sitting in storage.

Also as a side note, to get the best performance out of a f-theta lens you need to have the galvos placed a specific distance from the lens, to get a flat focus (otherwise the edges of your piece will be washed out). You also want to have a massive beam going into the lens, mine was designed for a ~1" beam going into the galvos.

[macona]

Thanks for the info, these are the two lenses I have, the silver one is about 4" in diameter. Ill have to call up melles griot and see if they have any info they can give me. Betting they wont say as it was made for someone else. Not sure of the focal length, looks like about 8" or so. The one on the right is for 532, I saw a seller on ebay with one and it is supposed to have a 50mm f/l

Business end:


f-theta scan lenses by macona, on Flickr

Galvo end:


f-theta scan lenses by macona, on Flickr

[Junktronix]

I've got a couple of ScanLab scanheads with Spectra Physics BL10 lasers (3W avg, but short 10ns pulses so >10KW peak power! which marks steel nicely) They are driven by an Scanlab RTC card. RTC connects to the scanners via high speed serial data (XY2-100 protocol) so it won't work with analog input scanners. I think some GSI scanheads also use this digital protocol.

I've been using an open source marking package called BeamConstruct - http://www.beamconstruct.com/. It claims to support EtherDream with a modified firmware.

[swamidog]

there were some issues with etherdream support. if i recall correctly, it was due to the beam construct folks not honoring the license agreements.

I've got a couple of ScanLab scanheads with Spectra Physics BL10 lasers (3W avg, but short 10ns pulses so >10KW peak power! which marks steel nicely) They are driven by an Scanlab RTC card. RTC connects to the scanners via high speed serial data (XY2-100 protocol) so it won't work with analog input scanners. I think some GSI scanheads also use this digital protocol.

I've been using an open source marking package called BeamConstruct - http://www.beamconstruct.com/. It claims to support EtherDream with a modified firmware.

[Junktronix]

That's too bad. It works nicely with the RTC card.

[macona]

Thanks for the info. That is just what I was looking for. I emailed the company that sells the ETH6608 card to see how much that thing sells for.

[Junktronix]

Let us know what you find out. I'd love to get my second marker going without buying another >$1000 RTC card!