DIY Laser cutter/engraver design advice

[picclock]

This is a project that's been in the making for some time. I wanted a cutter/engraver which would handle up to 1 Metre long items plus the usual stuff. Looked at all the std chinese and other stuff but was less than impressed by the quality, especially of the tubes, normally no name, and likely short life.

So probably because I was fed up with looking I decided (in a moment of madness  ) to use the DIY approach.
I need a working area 1 metre long by ideally 500mm wide, but I could shrink the width a bit if needed.

I've ordered most of the laser specific parts, and a Leci W1 tube, but will try to leave enough room for a 1200mm tube for futureproofing.

The Controller
Most people seem to rave about Lightburn, so compatibility is would seem desireable. Lightburn will allegedly work with Marlin, so at the moment I'm thinking of using one of the 32bit budget boards as in my 3D printer, SKR v1.4.
I can then use the trinamic drivers (2209) and I have confidence it will work, but I'm not sure if this is the best approach ..
The alternative is to up the expense and use a ruida controller, but its difficult to know if this would offer any useful added features, also does it support corexy ???.

The laser head, mirror and lens are very low mass so I am considering using a corexy type motion. But I have never used one and have no experience of them. Not sure how good it will be over such distances. Reason for trying it is to get rid of moving the  stepper weight with its associated speed and ringing issues.

For the frame vslot extrusion, 40mm square, 20mm square for x,  with wheels for bearings as per my Ender printer.
A stepper on each side of the Y, plus a small stepper for the X., assuming cartesian.

I've sort of worked out the layout in my head, tube at back top, with psu's (laser and controller/motion), air pump and other stuff underneath at the back. Control panel on lid with laser safety switch.

I'd be very happy to hear any thoughts about these issues, especially if you have experience of Marlin
and lasers.
 
Best Regards

picclock

[efrench]

How fast does the laser head need to move?  Have you considered LinuxCNC?

[picclock]

@efrench

Thanks for your interest.

I'm not sure how fast the movement needs to be. For engraving is likely the fastest use case, ignoring moves between firing, which can be as fast as possible.
 
I did look into linux cnc a while back but for this I'm thinking a dedicated controller which would integrate all the many needed functions (Fume extraction, water pump, nozzle air, laser power etc)  would be a better approach as its more compact, lower power consumption. I'm at the age where ease of use with minimum aggro is important to me 

Best Regards

picclock 

[BillTodd]

Following along with interest.

I have a couple of 400mm linear motion assemblies . I'm thinking a small router or laser cutter would be useful. Maybe i could combine the two functions?

Where did you buy the tube and mirrors?

[picclock]

Hi Bill
Depending on what you want to cut or engrave a diode laser may work for you. No mirrors, water cooling or high voltage power supply. Co2 lasers are big, typically 1M plus long. Also they can make lots of smoke, which means a dedicated enclosure with fume extraction. 

Only reason I am doing mine is that I couldn't find a readymade with the correct footprint. Mine will sit on a dedicated bit of kitchentop  , and I'll weld up a trolley with a shelf for the air pump/water cooler and reservoir. 

I also opted for a Ruida controller as it seemed the best fit.

Good luck with the router.

Best Regards

picclock

[MetalMagus]

Hello,
I am thinking of doing the same thing some time in the future. I "lurk" in several laser groups on FB and most use the Ruida controllers in combination with lightburn.

I came across the site below for a DIY laser build. It has downloadable plans and could be modified for a true home built affair without buying any hardware from them.

http://www.lasersaur.com/

regards

Sean

[picclock]

@MetalMagus
I've looked at several diy designs on the internet including the lasersaur one, but I have never been impressed. In my mind I keep seeing all the wrong engineering decisions. Look at the lasersaur. The way in which he uses aluminium section is like he has shares in the company, quite absurd. The y axis is driven by a belt reduction shaft which connects to two toothed belt pulleys which drive each end of the X axis rail. On mine I will drive the X rail with two stepper motors connected in parallel to a stepper driver, so much simpler and more elegant. I would mention the X axis motion but I grow weary. Check out the video

I used to lecture at a technology college and half of the students would try to copy this stuff off the internet and present it as there own work. All rather sad.

@BillTodd
I forgot to add I bought the laser (Reci W1) and mirrors from aliexpress. Brexit has altered things for them as they now collect the VAT for the UK (or it appears that way). So when you order stuff and pay for it 20% VAT is added**. Lasers have 0% import duty (Customs and Excise), so having paid my costs (£209.78) it should just turn up  , one day.

Best Regards

picclock 

** Apparently this only applies to goods of value up to £135. In my case for the laser (£146) I will probably have to pay VAT on top (~£29), and knowing carriers, a transaction fee on top of that.   

[MetalMagus]

I quite like the look of the guiderail sets from Cloud Ray. But they still use a drive shaft and a single stepper drive.

https://www.cloudraylaser.com/collections/guide-rail-set

One day I may get around to building one of my own

cheers

Sean

[picclock]

FxFrameExpert may help if you are designing extrusion assemblies. It will even calculate beam deflection so you do not have to overengineer your designs, and best of all its FREE.

http://maytec.com.de/index.php?id=21&L=1

No association with the company.

My Ruida 6445 controller arrived, but without software, manual or info. Cloudray sent me the links.

The manual for Ruida RDC6445G via link:
http://cloudray.oss-us-west-1.aliyuncs.com/Controller/Ruida/RDC6445G/User%27s%20Manual%20of%20RDC6445G%20V1.0.pdf

RDworks software.
http://cloudray.oss-us-west-1.aliyuncs.com/Controller/Ruida/RDWorksV8Setup8.01.48-20200418.rar

These are the V slot guides I am using.

Best regards

picclock

[WeldingRod]

My 3D printer design also used the drive shaft method.  You can see it just above my cats head.  It gives very positive synchronized timing on the two parallel rails, with a mechanical adjustment if you want it.  Dual steppers depends on you not accidentally bumping it for timing!

You should consider how the various bits are constrained also.  Ideally you want one end of the cross bar constrained differently than the other.   You dont want anything that binds up if the two parallel rails arent quite parallel.  I have pictures if you are interested in how I did it...

Sent from my SM-G892A using Tapatalk

[awemawson]

Very good resolution on that Cat model - did you draw it yourself 

[vtsteam]

The fur looks almost real, and the green LEDs were a nice touch! 

[picclock]

@WeldingRod
What an excellent cat picture, was it a scan ?

Using a rod to drive the two axis is a good solution to keeping things in step. My long shaft would be over 1M, and is probably a safe solution. However in my experience, two motors driven in parallel will not go out of sync during normal operation unless moved by hand. If it proves to be an issue I can use a fallback by triggering limit switches/sensors on each side to disable the stepper pulses for the relevant motor. The pulses would be re enabled by the change of direction line level. Failing that I may try your idea  . time will tell. I have seen one design on the internet which uses two steppers so clearly it works for someone.

My main concern is the X axis which at over 1M long may tend to resonate. The weight of the mirror lens assembly plus the drag of the air feed tube / led pointer wires makes it tricky to get a feel for. The motor end of the axis will be fixed to the gantry. The far end will allow for some change of length, possibly a shaft in the end of the extrusion into a linear bearing.

Thanks for the input

Best Regards

picclock

PS also thought about replacing the led on the laserhead with a couple of cheapo mW power visible lasers and setting them up such that the light dots merge when the correct focus distance is set (Dambusters style). that way you get accurate position and focus.




 

[WeldingRod]

I had problems with my dual Z screws getting bumped differently and tilting the works until I put a belt between them.  Of course, it ran smoothly enough that it would go down under gravity with the power off.  Belting them together stopped that problem too!

Sent from my SM-G892A using Tapatalk

[picclock]

My Laser tube finally arrived  . Its a Leci W1 tube. Scanning the barcode takes you to the leci site which says it checks out as genuine part. It has a performance graph on the tube and you can see that anything above 75W is optimistic

Did a functional test and it burnt a 5mm spot in some blue tape and wood. Nice round spot, so it all looks good. 

Working out the layout I'm definitely going for the driven shaft method for the Y axis. It means I can reposition the motor to get more useable depth.

Just need somewhere to store it safe until I'm ready to fit it now. Looks very delicate !!

Best Regards

picclock