Frankenstein Laser Engraver

FeaturedContest WinnerContest Winner

This Frankenstein Laser Cutter was built out of an old scanner and printer.
The whole thing evolved around the instructable of Groover and his 'Pocket Laser Engraver'.
This is a Making-Of. Although a lot, if not everything, of the mechnical construction requires ingenuity I tried to document the complete build process as much as I could. Every scanner and printer mechanics are different so this could not be used as a step-by-step guide. More of a "how it can be done"-guide. I try to cover the questions that could arise in the process of making.

I had absolutely no clue about electronics. All I knew was that RED is (often) + and BLACK is (often) Ground.

Therefore I have learned a lot in this project. Starting from mechanical stuff like self-replenishing brass bearings to electronical stuff like stepper motors and the difference between bi- and unipolar motors to soldering and etching my own board.

The work area is 270mm x 200mm. Just about right do cut some flip-flops for the summer.

It is able to cut :
  • foam rubber
  • tape
  • vinyl
  • paper (nearly every color but white/red)
  • several plastics (could require several runs depending on thickness)

It can engrave:
  • light wood (Balsawood/Poplar)
  • leather
  • bone
  • horn
  • plastics
  • some varnishes
  • blank cds/dvds (purple/blue)
The color is quite important. White and in general bright colors are hard if not impossible to work with.
Those colors reflect the red laser beam to much.
Red(ish) colors are a problem as well as they reflect all light in the red range spectrum.

The building costs (without mispurchase [easydriver clones were for the trash can]) is around 45-55 €.

Arduino(clone) 10 €
Easydriver x 2 20 €
Electronic bits and pieces 10-15€
Aixiz housing /w lens 6 €
Alu-profiles 5 €
------------------------------------- ----------
Total 46-56€

Well I forgot the Laser Safety Glasses (THIS IS A MUST!!):


Laser safety glasses 50 €
New total 100 €

Build time with knowledge acquisition and waiting for shipping was around 4 month...
I am constantly updating this so be sure to come back from time to time for further improvements.

May, 9th 2013 : Updated Step 14 : The Laser diode (pictures and some focus hints)
May, 13th 2013 : Added Step 17 - Links and files section
Added Step 18 - Take it to the next step (Improvement - optional)
Updated BOM list. Now contains more stuff you need
May, 20th 2013 : Corrected mm/sec to mm/min ! Seconds would be very very fast.

June, 1st 2013 : Added Step 13 - Alternate laser driver shield (Easylaser Shield)
December, 4th 2013: Updated Alternative Easylaser Shield schematic/layout with the help of jduffy54.

Remove these adsRemove these ads by Signing Up
elabz11 months ago
Nice project! The scanner base gives a very nice area to work with. I've dabbled with DVD-RW based laser engravers/cutters (a lot) and their 1.5"x1.5" (38mm x 38mm) workarea is quite a limitation - I'm sure having 270mm x 200mm is quite liberating!
What are the accuracies you are getting with this setup (96SPR + pulley) - what diameter is the driving pulley and how close are you able to get to the max resolution with EasyDriver?
ianmcmill (author)  elabz11 months ago
Having such a big work area is really quite nice !

I am getting an accuracy of 1/10mm (0,0039"). The x-axis driving pully (scanner) is 9,1mm in diameter(0,358"). The y-axis pully has 6,4mm diameter.
For the resolution, the scanner stepper with 96 steps does very well with 1/8 stepping. The printer stepper is running in 1/8 stepping as well but as it moves the motor doesn't sound very "clean". Actually it has a nice accurarcy of 1/10mm, too. But 1/4 stepping would be better I think. Another option would be to replace the stepper with another, bigger (in number of steps) stepper motor but I havn't found any good replacement (slavage) by now. I am learning something new everyday and things get clearer the more I read and try.
elabz ianmcmill11 months ago
I'm coming up with (9.1mm x 3.1415)/96 = ~0.29mm as the max resolution for the scanner motor with 9.1mm pulley in full step mode. I am curious if you're able to increase that by using 1/8th microsteps. Did you measure 0.1mm?

I've done such measurement before by cutting a line at a very acute angle. Say, you want to cut a line along the Y axis but have the X at the beginning to be 1mm off from X at the end. Doesn't matter, 1mm more or 1mm less, just has to be off. The longer the line the easier it will be to see the accuracy. Once you cut the line, you should be able to see quite clearly 10 saw-tooth like regions that break the straight line (hold it toward a light source to see easier)  if your accuracy is 0.1mm, 5 teeth if you accuracy is 0.2mm, 4 teeth if 0.25mm and so on.

I've done it for DVD-RW steppers with the stock 3mm pitch lead screw which in theory should have 0.15mm resolution (3mm/20SPR =  0.15mm per step) and I cannot see any improvement whatsoever in any of the microstepping modes. I've also played with the current setting of the EasyDriver - nothing I did could create any better resolution than what's available in full step mode. I chucked it up to my stepper being overloaded. I wonder if you get any different results from yours.

By the way, I actually found 0.15mm accuracy to be quite bad for my tiny cutter. The details I would want to cut were so small that I needed them to be extremely accurate and 0.15mm just didn't cut it (pardon the pun). See if you can cut yourself a tiny 1:212 scale  model of Lockheed P-38J Lightning from here out of 3mm craft foam - it only holds together without glue if the accuracy is better than approx 0.05mm. Same with the tiny rocket model there - if it holds together, it's at about the maximum resolution that you can get from this laser cutter. It will be really difficult to  focus the laser into a spot less than 50 micron anyway.

ianmcmill (author)  elabz11 months ago
Okay trying this one but I think the printer stepper is the crucial point or it's microstepping mode. I did the rocket you coded on your page and I needed some glue to hold it together.

you should be able to see quite clearly 10 saw-tooth like regions that break the straight line (hold it toward a light source to see easier)  if your accuracy is 0.1mm, 5 teeth if you accuracy is 0.2mm, 4 teeth if 0.25mm and so on.

You mean like "M3 X100 Y100". Cutting a diagonal line ?

elabz ianmcmill11 months ago
G1 X1 Y100

I think this would be the right Gcode as interpreted by Grubl (if M3 turns the laser on, M4 off). In other words, the X travel should be minimal, ideally 1 mm.

I don't know idiosyncrasies of Grubl's processing of the Gcode. Perhaps the construct you gave will work but even then it would read as
M3 X1 Y100

ianmcmill (author)  elabz11 months ago
I had a chat with elabz and here is the result of his Gcode from above

Top lines show the accuracy of the Y-axis.
Bottom, that of the x-axis. The x-axis is pretty clean so far. But the y-axis lacks accuracy for now. Switching the stepping mode could help. Or some dumpster diving.
ianmcmill (author)  ianmcmill11 months ago
ianmcmill (author)  ianmcmill11 months ago
(9.1mm x 3.1415)/96 = ~0.29mm as the max resolution

Ah okay now I understand what you mean by 'max resolution'. The more I think about it, the more my "measured" accuracy cracks. I measured it with a ruler that has 0.5 mm steps (0,01968 inches) and I just estimated this 0.1mm.
elliot42063 days ago
Okay thanks. And my other question?
ianmcmill (author)  elliot42063 days ago

Maybe you should read the whole instructable and have a look at the pictures. There is no generic "Part No. 123B".

elliot42064 days ago
Could you please explain exactly which parts we need from the printer? And can I use the laser from my printer?
ianmcmill (author)  elliot42064 days ago

Hey elliot4206

You need the laser diode (red light) from a DVD burner.

KrisCross931 month ago
Hi, I'd like to make the lasershield, but can't find the download for the fixed version. What exactly is wrong with your orignial layout?

Cheers Kris
ianmcmill (author)  KrisCross931 month ago

You can find a link to Groover's laser shield in Step 11.

Step 13 direct download link to my alternate laser driver with jumpers for different microsteppings.

ianmcmill (author)  ianmcmill1 month ago

"My" original layout was done with microsteppings in mind. Though I made a mistake where different jumper settings did not change the microstepping mode. Jduffy54 took the Eagle layout and fixed it.

If you like I can upload "my original" (non-working-jumper-layout) and practise your Eagle skills.

Hey, thx for your help. I finally found both download links, seems like I had to be logged in to see them...
But when I open the fixed one with Eagle it looks wrong to me? (pic)

rcerchier2 months ago

the jp1 would be have a jumper?

rcerchier2 months ago

hi, i make the circuit by Goover but when i put laser in the terminal it doesn't
work..If someone would help me on this problem my e-mail is

ianmcmill (author)  rcerchier2 months ago

Did you check the laser diode with a battery first ?

Are your soldering joints OK ?

Try to measure your circuit on continuity.

I bought this laser diode The joints are soldering ok..

the only different component is 6ke15a rather than sa15a

foto 1.JPGfoto 2.JPG
ianmcmill (author)  rcerchier2 months ago

Directly measure on the terminals if current is flowing.

ianmcmill write to me on these email please

ianmcmill (author)  rcerchier2 months ago

M04 for "laser on"; M05 for "laser off" in the g-code sender

ianmcmill (author)  rcerchier2 months ago

In the ebay description it says "These are multi-mode display diodes similar to the 1W+ 445nm diodes. Check these threads for info on output and beam correction."

I don't know if they are wired different then the DVD diodes or the
diode I used. Did you check if the diode is OKAY with a battery ?

the laser is ok, tomorrow at shool i measure on the terminals if current is flowing, with the command m04 the laser doesn't light

rcerchier2 months ago

hi, i make the circuit but when i put laser in the terminal it doesn't
work..If someone would help me on this problem my e-mail is

udaykhose3 months ago

sir can you please send me bipolar stepper motor driver circuit schematic insted of Easy driver which is compatible to this project because i didn't get Easydriver in our local electronics market.

Taj Mahal Wallpapers 1.jpg
kmyers84 months ago
ok first of all this is not beginner stuff lol this is sum hard shit even for me and I know a little bit about electronics so u must b a genius for putting this together. so major kudos to you my man. and on that note im gona ask for your assistance in making this cause im very interested in making one and would appreciate any help u may provide!
ianmcmill (author)  kmyers84 months ago
Thank you. Sure. Get cracking ! I'll try to help as much as I can.
davsvatos4 months ago
Wonderful instructions. I have one question. Can this procedure be applied to the stepper motors from DVD drive? Thank you.
ianmcmill (author)  davsvatos4 months ago
Yes, of course. They behave all the same.
Jeanot13144 months ago
Nice tutorial thanks!

I am working on a similar project, i killed already 3 printers to get my XY axes but on all of them it s not 4 wire step motors, it s a normal 2 wire electric motor! Do you think i could use this (maybe with a transistor to control) ? If those big printer brands succeed to use them they must be a way...

ianmcmill (author)  Jeanot13144 months ago
Good you point this out !
Newer printers hardly have stepper motors anymore. They use ordinary DC motors with encoders. You need an extra IC to control those motors. All in all it is very inconvenient to use. Although GRBL does not support this kind of motor controlling.

The printer manufacturer save a lot of money when replacing stepper motors with dc motors. I recently disassembled a HP multifunction printer with scanner and printer and there was not one single stepper motor. They even save on the belts and used toothed racks instead.

Therefore older printers are better to salvage stepper motors.
I spent some time today trying to figure out solutions... there is a circuit from Sparkfun based on a TB6612FNG who suppose to be able to drive DC motors in PWM at 100kHz. But i don't really want to wait a week for the delivery so i am gonna try with few transistor and a H-Bridge. I have no idea how how precise i can be with this kind of motors and fabrication (hope less than 500um!), will see ^^
Japaluca5 months ago
Hi, I'm new to this subject, not etiendo rather than for serving this step
fritzerik7 months ago
Hi Ian, thanks again for your answer! Your support is very helpful :)

I've found the source of the problem. When installing laserengraver from groover, this overwrites dxf_input.inx. So I reinstalled inkscape to get the original back.
Then, the z-depth in the orientation tools was not set correctly. I used 0, so it would not try to use the z-axis. But this resulted in the empty gcode files. So, I think this was the source of the problems.

But now before it makes the lines it tries to move the z-axis which is not there. Is there a way to stop it from doing this? The problem is that it already cuts when it should not. I'm engraving mirrored plexiglass so it is very sensitive to being burned too long.

Do you have a solution to this?

ianmcmill (author)  fritzerik7 months ago
The problem with cutting when it should not came up here as well. Only after I installed the bought laser driver from aixiz this problem was solved. I really don't know why the original laser driver from Groover had this problem but I could imagine that the relay could be the cause. But I am just guessing.

As for the z-axis. You could try to increase the z-axis speed in GRBL settings to something very high like 5000 and at the same time set the z-depth in gcodetools to 0.1 or something > 0. Anyways always do test runs on non-critical material like cardboard.

Engraving on mirrored plexiglass sounds interesting. I wonder if the red laser is capable of engraving it at all.
Hi Ian,

Sorry for the late reply. I've been busy studying... anyways.
I solved the problem that I had by using notepad and search/replace to remove everything that incorporates a z-axis. This works very good.
But, I get some strange behaviour from inkscape. When I convert the image to area filling, zig/zag, then it does not fill the black areas completely and I get some strange lines that do not correspond to the lines that should be (see image).
It seems as if the angle at which the zigzag goes (in this case 45 deg) gives problems at the object having 45 deg parts.

Have you ever encountered something like this? Would you have an idea what might be the problem here?

ianmcmill (author)  fritzerik7 months ago
I encoutered this issue, too. You can try to simplify the 's'. Path -> Simplify or Ctrl+L. This reduces the nodes of an object.

Get More Out of Instructables

Already have an Account?


PDF Downloads
As a Pro member, you will gain access to download any Instructable in the PDF format. You also have the ability to customize your PDF download.

Upgrade to Pro today!