On Wed, Apr 15, 2026 at 08:44:12PM +0100, Robert McWilliam wrote:
Hi,
Sorry this has ended up as quite a long message. This is the version
after I've done some fairly brutal editing :)
My current understanding of the hackspace laser(s):
We have a Trotec CO₂ laser that doesn't work. It goes through process
fine from SW point of view but doesn't actually seem to fire the
laser. It used to be a bit intermittent and if you rand it a few times
it would start working, but lately (for a maybe years long version of
"lately") has been giving nothing. A couple of folks have had a look
at diagnosing/fixing it but not really got anywhere.
We also have a K40 CO₂ laser (on loan, rather than actually being
hackspace property) that does work (or did when I last tried it, but
that was at our last location so before 2019?) but would need cooling
and extraction/filtration sorted to be able to use it in the space.
Is all of that correct? I'd be surprised if the years long passing on
of details between various folks or me misremembering hasn't gotten
some of it wrong.
I think the Trotec would be a good base to work from to frankenstein
together a working laser.
Extraction and filtration good enough to use a laser in our current
space without getting a visit from the fire brigade (or giving anyone
cancer from breathing the fumes) is likely to be trickiest bit to
solve for a DIY laser. That is already a solved problem for the Trotec
so we "just" need to bodge a working laser into it without breaking
that.
Building a fankenlaser will be a lot easier if we replace the Trotec
controller and SW with open source versions. I've been doing some
reading on what current options are (things have moved on a lot since
the days when GRBL was new that I last played around with this stuff).
I'm leaning towards FluidNC for controller and Rayforge to feed it.
But haven't actually used either so would be keen to hear from anyone
who has, or has other preferences.
http://wiki.fluidnc.com/en/home https://rayforge.org/
One of the FluidNC devs sells a variety of boards:
https://www.elecrow.com/store/BartDring
The other high level thing we'd need is a working laser to control.
It's possible the existing laser tube in the Trotec can be made to
work but I wouldn't rate that as very likely.
We might be able to take the working laser and HV power supply from
the K40 and transplant that into the Trotech. Unknowns there would be
if the K40s owner is OK with us gutting it for parts and how much
bodging would be required to make it fit to the Trotec.
If we want/need to get new parts I've seen some recommendations for
Cloudray. Tubes:
https://www.cloudraylaser.com/collections/glass-co2-laser-tube Power
supplies:
https://www.cloudraylaser.com/collections/power-supply
Matched bundles:
https://www.cloudraylaser.com/collections/tube-bundle
Another option would be to go for diode laser rather than CO₂. That
simplifies some things (no need for HV, water cooling or aligning
optical path through the machine, much longer lifespan than CO₂ tube)
but possibly complicates other things (can the existing gantry support
the weight? Is there enough Z clearance? Blue rather than IR changes
material compatibility. Might need to replace the window on the lid to
be safe with the different wavelength).
Random aliexpress listing for an idea of cost/options going diode
route:
https://www.aliexpress.com/item/1005007444269352.html
What I'd like to do now: buy one of the FluidNC controller boards and
hook that up to the Trotec to see that we can get it moving and
starting chiller, extraction+filter etc. Then (or in parallel) figure
out details of what would be needed for the different options for
actual laser to make a decision on that.
Anyone have opinions on that plan?
Anything now is better than nothing, I support you tearing a down and
rebuilding the laser cutter.
This is likely to go a lot faster if someone else can run with it
(it's been languishing on my TODO list for ages already and all I've
actually done is spend a couple of train journeys reading about
options). I'm definitely keen to help with this but if just left to me
it's likely to get lost among the may other things I'd like to do when
I have time and then get forgotten about for a few months.
Well done to anyone still reading at this point :)
It wasn't so bad.