CO2 laser - the learning curve

[monkeysarefun]

What printer?

 

Jim.

 

The Anycubic photon..

 

https://www.amazon.com/ANYCUBIC-Assembled-Innovation-Off-line-Printing/dp/B078N2TSYS

 

http://www.rmweb.co.uk/community/index.php?/topic/133638-tests-with-my-new-3d-printer/

[Simond]

Simon,

 

Good news that you had a successful replacement.

 

I note from your tests for number of cuts to get a cut through that you use the same as I used to - i.e. cutting squares of (I guess) 10mm side length.   Do you find that the results from these tests are slightly under when you cut much longer perimeters.  I have found on my 4W Emblaser 1 that I usually have to increase the number of cuts by at least one if I have done the test with 10mm squares.   I wondered if it was the MDF retaining some heat with the shorter perimeter and improving the effectiveness of succeeding cuts.

 

Jim.

Hi Jim,

 

Thanks for your comments. The squares are indeed 10mm, but only one pass. That’s the big difference with CO2, it’s nominally 40W.

 

Monkeys,

 

Stop it,

 

I haven’t got room for 3DP, as well as everything else. And MrsD may lose her sense of humour...

 

Best

Simon

[Simond]

Loco shed trial assembly pix now on the Porth Dinllaen thread

 

Best

Simon

[flubrush]

Hi Jim,

 

Thanks for your comments. The squares are indeed 10mm, but only one pass. That’s the big difference with CO2, it’s nominally 40W.

 

 

Simon,

 

Sorry,  I had misread some of your earlier messages when I thought you were doing multiple passes.  I envy your 40W capability.  Multiple passes cutting out large pieces can be a long process on a 4W diode. :-)

 

Jim.

[flubrush]

The Anycubic photon..

 

https://www.amazon.com/ANYCUBIC-Assembled-Innovation-Off-line-Printing/dp/B078N2TSYS

 

http://www.rmweb.co.uk/community/index.php?/topic/133638-tests-with-my-new-3d-printer/

 

Apologies to Simon for the thread drift and thanks for your response.  I'm thinking of getting a Phrozen Shuffle which is a bit more expensive version of the same idea.  It has had good reviews and there is a good distributor in the UK who uses the machine himself.

 

https://www.bluecastjewellery.com/phrozen-shuffle/

 

Jim.

[Simond]

No apology required, it’s all grist to the mill, and, in these days of computer aided modelling, very relevant.

 

Do kick off a thread (and link it here!) if you get one. I’m tempted, but there are only so many hours in the day!

 

Best

Simon

[Simond]

Not strictly lasers, but related, a new toy arrived from China this morning...

 

 

Adequately packed, very thorough in fact.

 

There are no instructions...

 

This will be fun!

[flubrush]

A small CNC router??   :-)

 

Jim

 

PS By the way,  got some 4D 1.5mm MDF recently and I'm looking at trying it out soon and comparing with the Maple Street MDF I Have used up till now.  jim.

[Giles]

<p>well well..... It will be interstin to see how this goes!</p>

[Simond]

Yes, £115, including postage. GRBL controller, power supply, three steppers and more Meccano than you can throw a stick at.

 

 

Got to here, pretty much hassle free, but there’s an issue with the Y axis

 

 

I guess you can see that the rods are not parallel. This is because the plastic mouldings into which the ball slides (yes, real ball slides, with oil seals, proper stuff!) appear to have a draft angle on the mounting face. This is pretty stupid, as it means that the thing cannot work as delivered. Indeed, tightening the screws pretty much locks up the rods.

 

The option for today is the sand the tops of the mouldings to get an acceptable fit, and then shim it with slices of business card until it’s flat, level, and moves smoothly.

 

The good solution would be to 3D print some new slide holders, and arrange that they are adjustable. Maybe, maybe next week.

 

Btw, the motor holder and Z frame assembly is all 3DP, as are the bearing housings for the tails of the leadscrews.

 

The slide rods are ground 10mm bar, the leadscrews are 8mm buttress thread.

 

I’m impressed, and depressed in roughly equal measure. I couldn’t buy the material from wholesalers to make this for what I’ve paid for the kit of finished parts.

 

More later

Best

Simon

[Simond]

You can see the angle here - all four pedestals on the same rod. The scribble is black felt tip so I can identify where it isn’t touching.

 

 

This is after a couple of minutes’ work. Another 10 minutes or so and we should be rockin’ & rollin’...

 

Best

Simon

[Simond]

Sanding worked. Ideal job to do whilst on the phone to mum.

 

All assembled, and all axes seem to move smoothly.

 

Next step will be powering it up.

 

 

Quite a few bits left over...

 

 

Well, not really. 4 T nuts and Allen screws, set of Allen keys, two spare lead nuts, and a spring. A bag of clamps and T bolts presumably for work holding, a collect and two milling cutters, power supply, US mains lead and very naf 3-pin adaptor, a CD and a USB lead.

 

I guess the lead nuts are supposed to be wound into the leadscrews with the spring compressed, to minimise backlash. There’s only one spring, so I may need to source another. There’s certainly a load of backlash in Y, so that needs sorting. X & Z do not seem so bad.

 

A pleasant diversion for the day.

 

More soon.

Simon

 

Oh, yes, the connection to lasers? As I said, there are no instructions for building it. But there is a single page entitled “How to Set the Moving Direction of the Axis”, which refers to two other devices presumably supplied by the same outfit, a “DIY8000 Corexy 1818 Desktop laser” and a “DIY 1720/4050 Desktop laser”. There are pretty little pictures of the devices, they appear to be solid state lasers with no enclosure. Might be cheap, however.

 

I guess you could fit a SS laser to this machine, but the bed size is too small to be useful, I feel. The bed is 160 x 100, the spindle appears to be able to reach almost all of it.

 

More later,

Best

Simon

[bgman]

Hi Simon, 

 

That looks like a very interesting piece of equipment !

 

Look forward to seeing what you produce.

 

G

[Simond]

DISASTER!!!

 

Got it working, it all ran correctly, correct directions first time, very pleased. Was manually scribing in a bit of MDF, using the provided software. Then I hit the “retract to safe Z” button, which hammered the tool into the workpiece until everything stalled. Safe z is up, not down... apparently this is a known issue!

 

And then it wouldn’t run. Nix, nuffink, nada. Dead as the proverbial parrot.

 

So I got the multimeter out, and the power supply was dead. That’s a surprise, assumed a board failure, but power supplies, I can do. Before opening it to see what had died, I unplugged and plugged it in again. Solved! Looks like it has a lock-out.

 

Now I’m into dxf2gcode. This is currently being a pain and importing DXFs as if they were in inches, despite every setting I’ve so far located being in mm. I’ll find it...

 

More soon...

Best

Simon

[raymw]

g20 to g21, maybe

Edited December 2, 2018 by raymw

[Simond]

Thanks Ray,

 

Yes but, no but...

 

It appears that D2G imports the DXF in inches, converts it in inches and outputs gcode, in inches. This would be great if I drew my CAD in inches, but I’m not Imperial.

 

The CAD models are all metric, DXFs are essentially dimensionless, but do have a header in which units can be specified. I’m using a rather old version of TurboCAD, and I suspect it defaults to inches, and so labels the header. It does a couple of other weird things like watermarking the file, which completely b%%%%s up the autoscaling on everything downstream, and which I now use LibreCAD to eliminate - it’s a tedious workaround until I get a new laptop, with some newer software on it.

 

(Well, actually, I had, young MasterD decided he needed a new laptop for university, so I agreed to get him one, and rebuilt his cast off whilst waiting for him to choose something. He then decided that the rebuilt one was fine, and took it. “thanks Dad!” Bah!)

 

It appears that the problem is known, there’s a bit on the www about it, and my first port of call will be a reinstall of D2G.

 

Will post up results in due course.

 

 

Digging a bit further in the machine cd, I discovered that there are instructions, and they’re not at all bad, neither is the control software, other than the incorrect z command. Apparently they do three sizes of the machine, the largest will certainly do A4, and there is a laser “head” available too.

 

Best

Simon

[raymw]

Hi Simon,

I installed d2g, last night, it all displays as mm for me, but I never got very far in getting actual code from a dxf file (a bit sleepy). afaik, I don't have inches/mm set in my dxf files, not checked headers because I've never had a problem with what I do. If you can, maybe you can save the dxf file as another version. dxf can be a real pita.

[Simond]

Still playing with it. Everything, including LibreCAD is importing in inches.

 

I’ve even used GEDIT to edit the DXF file and specify “4” for mm in both INSUNITS and MEASUREMENT, and it still imports in inches.

 

Either I’m being very dim, or there’s a very well hidden setting somewhere!

 

DXFs are indeed a PITA.

 

There’s always the possibility of writing the gcode... no5 too bad if you just want to drill some holes, but any profiling would be an exercise in concentration.

 

Best

Simon

 

Could I beg a favour? Would you post a small metric DXF file?

Edited December 2, 2018 by Simond

[raymw]

Hi Simon,

 

Hopefully attached is a dxf file of a simple 2d drawing of a bearing housing. It has a dimension line, and a triangle and a number of circles. If you open it in Irfan View, with it's plugin, it gets the wrong size text for dimension (btw, the dimension is not to the circle centres,) If I open it in my cam software, it does not show the dimension at all. The dxf version is R12, and I've set no measuring mm/ft within my cad program. I always draw full size, with undefined units - it allows easy scaling later on, if needed. Most of the milling I do is more or less 2.5D. I draw plan view, then select areas to be pocketed and lines to be profiled etc., within the cam software. I've never needed to draw anything in 3D. I have not used the software you are using, other than a quick test last night. It's a good job your crash was not on a more expensive/powerful machine. Not sure of the usefulness of a 'retract to safe z button' would be. I hope it's not an alternative to an emergency stop. If something is going wrong that you need it to stop doing it asap, not automatically move anything else.  I've attached the g code file produced from the dxf. It may not be of much use to you, but could let you mentally reverse engineer the g-code to get an idea of what the dxf implied z values could be. I've got some basic, very basic, g-code instructions on my old web site at cnc.yertiz.com and a free engraving program, if you want to play, but I expect you'll need to mess with the post processing, but maybe not.

 

Best wishes,

 

Ray

colinbearing.dxf

colinbearing.txt

Edited December 2, 2018 by raymw

[Simond]

Ray,

 

Thanks! That’s very kind. I’ve been fitting limit switches today, and getting them working. There is no big red button! (Yet)

 

I know all about crashes on larger more expensive macghines - last year I stuffed a several grand Renishaw probe into the workpiece & vice of our then-brand-new Doosan 4500. Happily, it was only the probe that suffered (that, & my ego, confidence, etc.)

 

I’ll have a play and let you know how I get on. The units thing is very strange.

 

Thanks again

Simon

Edited December 2, 2018 by Simond

[raymw]

Limit switches, on home diy machines, can be problematic. I don't fit them. They normally operate at ttl logic level, or less, and can be a good source of getting 'interference' into the controller. You'll need screened leads for them, and screening connected at one end only, and a single star point for all your earths, - logic, motor, drivers, mains, else you're likely to get spurious earth currents, that can give false triggering. If you want, you could send me your dxf file, and your g-code, I'll see if I can spot anything. (fwiw, I had to change the .dnc file extension to .txt  as rmweb complained)

Best wishes,

 

Ray

[Simond]

Thanks Ray,

 

If they’re a source of trouble, I’ll simply unplug them, I rather thought that having a home position worthwhile. I haven’t had issues with them on the laser. The earth points seem to be a single busbars along the back of the control board. Fingers crossed.

 

The DXF was simply a “Japanese Flag”, a circle in a rectangle, nothing more. I’ll pop the file over when I’m on the laptop.

 

Best

Simon

[michl080]

Limit switches, on home diy machines, can be problematic. I don't fit them. They normally operate at ttl logic level, or less, and can be a good source of getting 'interference' into the controller. You'll need screened leads for them, and screening connected at one end only, and a single star point for all your earths, - logic, motor, drivers, mains, else you're likely to get spurious earth currents, that can give false triggering.

Ray,

removing safety switches is not the best idea.

You may want to "preload" the switch input to avoid erratic signals.

a resistor pulling up the controller input will normally help. The resistor value is not critical. Anything between 1k and 10k should be working.

 

This will only work if the switches are connected to ground. Lowside switching, where the switch is connected to the positive supply, need the resistor connected to ground.

 

best regards,

Michael

[raymw]

Hi Michael,

Thanks for your diagram. However, i did not mention anything about removing safety switches, I did say that I did not fit limit switches. For my machines, and for me, limit switches are not necessary. For many folk going the diy cnc route, they want to add as many bells and whistles as possible, which is fine if that is what they want, but I've found that often, later on, when they hit problems, there are many more things to check, and most likely they've lost their wiring diagrams, if they ever made any. Also, most do not fit a limit switch on the most 'dangerous' axis, (Z), and still end up milling into the table. For many of the retrofitted manual mills, the steppers are two weak to strain the bearings/screws/whatever, and will simply misstep if the table hits the end stops. If it is necessary to continue machining the workpiece, it will need repositioning, depending on how quick you do a manual e-stop, then you may be able to re-use the piece, maybe a better chance if you've hit a limit switch, since all axis should stop moving. But, why run g-code that is going to run to limit switch positions? Of course, for other machines, industrial units with far more hp, different operators/programmers, then limit switches will be essential, but it is only recently that Haas, for example, https://www.youtube.com/watch?v=JqfnlPELl_g&t=2s have sorted out the z axis programming error crashes. For different diy machines, with the lighter weight, 3d printer type mechanism, it may be that the steppers are strong enough to damage the framework, and if using software that maybe behaves badly, then limit switches may be essential - but, in reality, I don't think you should rely on them. Easy enough to check through your program before running it for real.

 

Best wishes,

 

Ray

 

edit to add (I have no connection whatever with Haas, or any other machine manufacturer)

Edited December 3, 2018 by raymw

[Simond]

"But, why run g-code that is going to run to limit switch positions?"

 

three points here, I think - it's probably easy to make mistakes until one has gained some confidence and experience with the system, so hopefully, they prevent some damage - even if the steppers will just misstep, they are used for homing, which is useful, if it is consistent, and more concerningly, the Woodpecker controller has a $-code for the max travel in each axis, so if homed, it "should not" exceed the max travel...  but it does     (I think there may be an $-code setting for that too)

 

There are no safety switches on this little beast at all - no guarding, nothing.  You'd have to try to mill yourself, but it wouldn't be difficult to get swarf in your eye.

 

Once I've worked out the best way to do it, I will fit an e-stop, ideally stopping all the feeds without losing position.

 

best

Simon