Setting the focus

[tele]

Recently I've bought a new black head with air assist to replace the original one. Its size is different, it sets the position of the lens about 10mm higher than the old one, so I have to set the table and recalibrate the focus.

This is my problem, I've found descriptions about calibrating mirrors, but nothing about calibrating laser focus.

What is the correct procedure ? Let's assume that I'm able to set the height of the worktable .

Where do I have to focus to ? At the surface of the object ? Or inside the object , somewhere at the middle ? How can I check when the beam is in focus , it has the narrowest diameter ?

Some people here in the forum said that the beam is about 0.1 ... 0.3 mm wide . I have crap eyes, I can not make the difference, " wow this is 0.1mm wide beam not 0.15mm ... wohooo !!! " - I can't do this with my naked eyes .

Please help, thank you so much.

[Techgraphix]

My table can be adjusted in height, but this procedure will also work if you adjust it while extending or shortening the laserhead's tube..
For autofocus i placed a proximitysensor just near the head. I use a metal block to autofocus and trigger the proxi..
What you need to make is a block with plane with a fixed height of, say, 10mm heigh and a plane with a slanted height from 5 - 15 mm (in height) and say 10cm long..
The middle of this slanted plane is at 10mm. Now you (auto)focus on the 10mm plane. Next you stick a labelsticker on the slanted plane and let the laser draw a line of 10cm long (low power please) . On the sticker is a line drawn that widens out on both sides.. Where the line is the thinnest is your focuspoint..
If that is, as an example, 2 cm from the middle on the lower side you know that your focuspoint is 2 mm lower than where you focus was set..
So you lower the focussensor by two mm and do the test again. If you did it right, the thinnest part of the line will be in the middle of the slanted plane where it is exactly the same height as the horizontal plane..
Here a picture of my own simple focus-finding-tool... I only use it when i want to adjust my proximity-sensor after changing the lens...

Focus-finder.JPG (56.99 KiB) Viewed 3488 times

Kees

[tele]

Thank you Kees. Not bad at all, I think this will do. Nice and simple, this is what I wanted.

Is this your idea, or its an official way and this is how the big laser guys do the things ?

Its definitely worth a try. Your focus is absolutely accurate then ? Any advices ?

So you put the focus on top surface of the material ? Is that good ? What about thick plywoods ? Can you cut thick things when you are focusing on the surface ?

That proximity sensor with metal block also sounds very interesting . Can we have some more info about it ? I didn't know that proximity sensors are so precise .

Although I guess I need worktable not just adjustable, but movable by z-axis, with stepper motors. Am I right ?

Thank you again, you saved me.

[Techgraphix]

I don't know if other people do or did it like this but to me it was a logical solution to this problem.. (i'm a technical mathematician for hobby )
My proximity sensor has an repeatability accyracy of less than 0.02 mm.. far better than a microswitch and easier to use..
I have a little film on youtube where i replace a lens. there you can see the use of the proxy.
https://www.youtube.com/watch?v=EnaLNZPRrwo
Don't notice the controller.. that was my first and i'm sure Marco doesn't like it (you know the story perhaps... )

Depending on your own interpretation you can determine the focaldistance within 0.2mm.. Measuring the distance between te lens and the object won't work. A 38.1 mm lens can have a focal distance between 30 and 45 mm!!

With the focalpoint on top of the material i cut 6mm acrylic with a 38.1 lens... I know 100mm would be better because of the Depth of Field (DOF).. But a bigger DOF also gives you less dot-density (a bigger focalpoint) that's why short lenses are used for engraving and large for cutting..

Real thick material: Or i set my focalpoint about 3-4 mm in the material or i use multi-pass wher i lift the table a few mm after each pass (U-axis + some electronics)

Kees

[tele]

I watched the video. Nice. You have much better laser than I have. Mine is K40 the small CCC (Cheap Chinese Crap) laser. Most probably only your laser head and z-axis driven workbench cost more than my whole machine. And I won't upgrade it (I'have bought honecomb plus air assisted head for it and that will be all ), because in that case I would have CC (Chinese Crap) laser, but not cheap anymore. I don't understand people who upgrade it. Why don't they buy a better one at the 1st step. Buying a cheap one then patching it is a silly thing in my opinion.

>> "Don't notice the controller.. that was my first and i'm sure Marco doesn't like it (you know the story perhaps... )"

Why ? Is it a crap controller ? (and I donno the story)
Imagine my controller. It has no control buttons and display, but analog ampmeter and a potmeter. I don't understand this controller upgrading thing neither. People are upgrading this crap K40 with AWC680 controller. I also was thinking about it. But actually what would I profit of that new controller ? I can't see any honest arguments or answers in the "Why upgrade to AWC608 ?" topic. My machine stage , rails, stepmotors, laser, power supply, laserhead, worktable, everything is the same. A controller alone can't make miracles. I can't see reason to upgrade. What will be so much better ? ( whats worth $600 ) Accuracy ? I don't think so. And still you have the same problem, you have an undocumented Chinese SW and HW. From a bucket into kibble. Only some old deep-sea wizzard knows these Chinese things. I like the well documented and specified things. You are a mathematician, you must understand, what I'm talking about.
I wish I could find an American or German or Japanese well documented controller with a good (and not Google translated!!) SW.

>> "Measuring the distance between the lens and the object won't work"

Not absolute distance measuring, but comparative measuring needed. You set the focus once, then you can make a spacer gauge for that distance, and you can set the distance between your lens and the surface with that, anytime later. I think its working.

Thank you very much Kees, I have learnt many things from you, I'm gonna build that focus slope, its a good idea

[Techgraphix]

I think the AWC608 is worth every penny of it.. It is easy to use and, indeed, the manual needs some attention..
I haven't found an affordable, well performing controller without dongle from any other country.
With the software comes a languagepack. you can adapt it, or even make it in klingon if you like. It's a plain textfile.

LanguagePack.jpg (5.18 KiB) Viewed 3463 times

Then the balance between this controller and the K40.. you're right.. But i think many people buy this K40 as a step-up.
I designed and build my own laser but that doesn't mean it is cheaper than any commercial 4-5K laser.
Not everyone has the knowledge to design machines so buying one for under 1k looks tempting.. But as you know, you can't build a qualitymachine for 1k
I won't call it Crap, just build with low priced material.. Nice to play with and learn somthing about lasering... before you step up or build a better machine.
In that case you can use the controller again..

Kees

[tele]

>> "...With the software comes a languagepack..."

No, its not about language pack. Its translated literally, and you have no idea, what does it mean (a menu item, or a parameter) in the given context. What will be the effect if you modify that parameter ?
You have to try everything, and learn everything. What happens if you push the button "duke" ? Or its maybe "puke" ? They wanted to say "nuke" ? Try it then you will see !
Documentation would be necessary in my opinion.

>> "...I think the AWC608 is worth every penny of it..."

I think the same. I have an impression that AWC608 is made for a higher class level Chinese laser cutter. (not for $500 class, but for $1500 class) No cheating there, Its worth the money, not bad controller at all.
But why upgrading ? Can't you decide how much money you have, and what do you want, what can you afford ? You buy a dog-cart, then you change the wooden bench to a soft velvet padded chair, and you think you have a Boeing-747 now.
I'm still thinking about upgrading. But I still can't see the rational reason. What will be better ? Precisely. Be specific. What will I get for my $600 ? I won't get answers because we are still in the field of the undocumented, unspecified, unwarranted things (Chinese) . "...everything is so much better..." , "...I have a powerful machine now..", "it was worth every penny"...so on, these are not so good an specific arguments.

I ask something : what is the positional accuracy of your machine ? This is just a mechanical accuracy without the optical one, and the controlling accuracy.
And people will not know this, not at all !!! And actually they don't even care. This is one of the most important parameter of a CNC machine !!! Isn't silly if we don't know the most important parameters, but we start to upgrade the machine ?
I don't buy eye glasses for my golden hamster. Poor creature can see no farther than 20'' , but I can't help on that, this is what she got from the God.

[tele]

One more thing Kees, I almost forgot !!
Would you please send me the technical description of the proximity sensor solution. I think I'm gonna build that, if its so accurate as 0.02mm .
If it can be public, of course. Its much better solution than measuring the distance.
Thank you !

[Techgraphix]

Proximitysensors come in various types but the type that suits best are the cylindrical types. a NPN type switches to GND on trigger.
These have 3 wires: Black = output, Brown = +V and Blue = GND.
Mines got a diameter of 12mm but they also exist in 8mm. These are lighter and as they move with the head, all weightreduction is better.
Try to find one with a triggerdistance of 2mm. these have a better repeatability than the 5mm types (0.02 for the 2mm and about 0,05 for the 5mm types)
There are also 2 types: flush and nonflush. You can recognize them on the "nose" Flushtypes only look "forward" and have the metal up to the sensing surface. Nonflush type also look a bit to the sides. The metal part stops about 4-5mm from the end:
Like this:

A nice proxy for the head would be a LJ8A3-1-Z/BX (1mm sensing) or a LJ8A3-2-Z/BX (2mm sensing)
The sensingdistance is for iron.. If you use Aluminium or Copper instead, the sensingdistance is much smaller..Don't know if this influences the repeatability..

If you like to have a drawing of my bracket, i will try to find it between the ~ 10.000 thing i made...

Kees

[tele]

>> "...If you like to have a drawing of my bracket, i will try to find it between the ~ 10.000 thing i made..."

It won't be necessary, this is enough thank you very much.