Laser tube

[Ernie Martinez]

I've seen some large format laser cutters, and was wondering if I could retrofit it to my cnc router. Does anyone know of a source for laser tubes which can be mounted in place of the spindle? I'm just talking form factor here.

[Martin Boekers]

I think it maybe much more than mounting on a spindle, I'm not sure a CNC table can move as fast as a laser.
Dan Hintz would be the one too know!

[Lee DeRaud]

I think it maybe much more than mounting on a spindle, I'm not sure a CNC table can move as fast as a laser.

CNC machines will move fast enough for vector cutting, but would be painfully slow for rastering: CNC router speeds are in the inches-per-minute range rather than inches-per-second. The problem is, a laser tube big enough to be useful for vector cutting is way too big to mount on the spindle of a CLC machine. (Reference point: the tube on my smallish 25W laser is about 6"x6"x16".)

There are examples of people doing this over on CNCzone, but they're using fractional-watt lasers and either cutting paper or taking all day to raster something. IMHO, it falls into the "it can be done, but it's not worth the trouble" category.

[Ernie Martinez]

I've seen a few where the laser looked self contained as opposed to flying optics Like we have in our engraving machines. A YAG laser comes to mind. Another alternative would be a fiber laser. Would just need to route the fiber optic cabLe to a lens assembly on the spindle mount.

[Ernie Martinez]

I just want to vector cut, but after some thought, it's not as easy as just bolting it on. I would have to be able to turn it on and off, and vary the power which would mean creating a new post processor, and using a pc based controller like Mach 3 instead of the dedicated controller I currently have. Can be done but would be a bit of work.

With respect to cnc speed, look at a Kern laser. Pretty much a cnc machine but has flying optics in the gantry.

[Tony Lenkic]

Ernie,

It sounds like you would like something similar to Kern laser system.
I don't know their approach to deliver laser beam and where laser source is actually mounted.
Their lasers are large format open bed like typical routers design.
Eurolaser is another system with similar setup.

[Paul Phillips]

FWIW, MultiCam makes a large format one that will raster up to 1800" min, not sure how that compares to my universal laser though.
http://www.multicam.com/eng/Products/2000laser.html
Paul

[Lee DeRaud]

FWIW, MultiCam makes a large format one that will raster up to 1800" min, not sure how that compares to my universal laser though.

1800IPM is 30IPS...not too bad: your laser is probably in the 50IPS range.

I'm not sure I want to know the price of that MultiCam unit, as too many zeros in a row tends to give me a headache.

[Dan Hintz]

CNC machines will move fast enough for vector cutting, but would be painfully slow for rastering: CNC router speeds are in the inches-per-minute range rather than inches-per-second. The problem is, a laser tube big enough to be useful for vector cutting is way too big to mount on the spindle of a CLC machine. (Reference point: the tube on my smallish 25W laser is about 6"x6"x16".)

There are examples of people doing this over on CNCzone, but they're using fractional-watt lasers and either cutting paper or taking all day to raster something. IMHO, it falls into the "it can be done, but it's not worth the trouble" category.

Listen to Lee... he said exactly what I would have, particularly the last part.

Another alternative would be a fiber laser. Would just need to route the fiber optic cabe to a lens assembly on the spindle mount.

This would be better in terms of weight, but you're still slow on the speed (you're still dragging around that heavy gantry).

FWIW, MultiCam makes a large format one that will raster up to 1800" min, not sure how that compares to my universal laser though.
http://www.multicam.com/eng/Products/2000laser.html

30IPS isn't bad, but when you consider even the Epilog and ULS machines are in the 75IPS range (and the Trotecs in the 140IPS range), it's a relative snail's pace.

[matthew knott]

How do these speed figures relate to real world speeds, does the engraving head ever actually hit these speeds? 140ips is 3.6 metres/sec! On a speedy100 that's means it covers the full length of its engraving area 6 times in 1 second! Back and forth 3 times! I can't see any videos where they get anything like close to that, so is the figure a real speed or theoretical maximum?

[Dan Hintz]

See this video at about the 0:40 mark...
http://www.youtube.com/watch?v=_u2esCreLXE

[David Rust]

Matt,
i'm no expert, however there must be an acceleration/deceleration curve that is programmed in. I am sure the manufacturers publish the fastest speed the head will travel. I was playing with my Epilog and was varying the speed on a job to figure out where the acceleration/deceleration point was. This is what I did... I was running a job of 65 pieces (water bottles on the rotary tool) with about a 6 inch wide x-axis raster. I know that the cycle of one sweep should be the following: accelerate from zero speed to programmed speed, maintain programmed speed, decelerate to zero speed, then repeat in the opposite direction. I started at 100% speed and noted the total time it took to run one bottle job. If i choose a speed too fast for the acceleration/deceleration curves then i will never reach the desired speed, i.e. i will be accelerating then immediately decelerating to meet the "reverse direction point". For each bottle after that I reduced the speed by 5% and noted the time to complete the job. I know that at some point I will hit my programmed speed in between the acceleration and the deceleration curve. I'll know that I hit this point when my total job time increases. As I stepped down my programmed speed I noticed an increase in total job time at about 70% speed. So, any speed setting less than 70% I reach my max head speed, "x" inches per second, for an amount of time before I start to decelerate to zero.

An acceleration/deceleration curve is necessary to smoothly start and stop the laser head, if this was not used the head will "jerk" at each end point and cause excessive wear and tear on the bearings, belts, motors, etc... Also a "jerk" in the system will cause a vibration/movement in the system that may induce movement on the engraved piece.

The acceleration/deceleration curves may not be fixed, they may be a function of the programmed speed. I am sure each manufacturer has there own proprietary algorithms in their micro controllers to minimize job times, wear on the motors and drivetrain.

My 2¢,
Dave

[Lee DeRaud]

Acceleration is indeed a large factor to be considered. And while for a purpose-built laser the moving parts in raster mode weigh only a few ounces, even on a small CNC router they typically weigh over ten pounds.

[matthew knott]

Im calling BS on that video Dan, i would stake money thats been speed up, infact not even speed up, its fake, and heres why (i could be wrong here, just imo) each x scan produces a readable line of text, how is this posible?? you would need a type writer to do this , you can create a character (unless its a '.') with one scan. We have a fibre laser with the fasted head ScanLab make, and its nothing like that quick. I think thats a misleading video and should have a disclaimer (unless I'm wrong) Im sure the Trotec users will be along to confirm or deny is this is real or not!!

[Lee DeRaud]

Im calling BS on that video Dan, i would stake money thats been speed up, infact not even speed up, its fake, and heres why (i could be wrong here, just imo) each x scan produces a readable line of text, how is this posible??

Speeded up? Possibly. But I saw multiple passes for each line of text. (Not many, it's pretty small text.)