A Video Diary: How to Use a Laser Cutter Tutorial Series 03

• Looking at dithered picture as opposed to a greyscale picture
• How does the laser work with both types of pictures?
• Design a special coarse resolution picture and set it to 64pixels per inch
• Calculate the ideal line spacing (interval) for this pitch.
• Test the pattern at several incorrect pitches to observe effects as well as the perfect pitch.
• Now take into consideration the spot size capability of the lens you are using
• Calculate the best dot to dot resolution you can achieve with your lens.
• Draw a new test pattern to match the resolution of the lens
• Recalculate a new interval to suit.
• Stressing the importance of understanding what the smallest spot size your lens is capable of achieving.
• Choice of material influences the effective spot size
• How to use a laser cutter tutorial

• Demonstration of how I go about checking the REAL rather than the claimed beam diameter.
• IMPORTANT. Remove your lens before carrying out this test
• Start off using mid power about 40 watts and burn to a maximum hole.
• Set to full power and the burn gets even bigger, indicating that the beam size is not a constant feature, but power related.
• It was necessary to open the lens down tube to accommodate my beam diameter.
• Examining the power profile within the beam by burning into a piece of thick acrylic
• Discussion of the beam power profile
• Look at the result of the “mode burn” test

• An unanswered question about laser mirrors remained. Real gold is supposed to be naturally 99% reflective.
• A correspondent has kindly volunteered to coat a set of my lapped copper mirrors with 24K gold so that we can do the FINAL setoff mirror tests.
• The results were showing about 97% efficiency which is on a par with my polished raw copper mirrors.
• Final recommendation for mirrors is….. buy molybdenum. They are not very expensive; they are very durable and about 95% efficient.

• Exploring Dithered graphics.
• There is a difference between ppi and dpi
• First you must prepare your picture Photo) n an external program. I use Photoshop
• Convert coloured pictures to grey scale and adjust tonal range.
• Change the image to a DITHERED bitmap where the picture is composed purely of black dots of varying density This density variation fools the eye and brain into seeing a grey scale image.
• Set the scan INTERVAL to match the picture resolution and set scanning parameters
• I use white card as a nice simple test medium.
• This is a 600pixels per inch resolution and the scan interval was also set to 600 lines per inch.
• Run test at various powers and speeds and examine results.
• What can be causing the 3D effects because this looks like some sort of result you may get from grey scale engraving.
• Grey scale engraving can only occur if you set the max ad min powers to different values. These results were achieved with max and min equal.
• The dithered bitmap was run on an inkjet printer to show that the picture was indeed a series of varying density black dots.
• Now we go into a very detailed analysis of the pictures, finding that we do NOT have a pattern of uniformly coloured dots., but many shades of brown dots and lines.
• Now follows a long step by step analysis of what may be causing varying power dots
• Conclusion is that it is all to do with the response time of the beam forming and decaying
• These findings are a steppingstone into a whole new set of investigations.
• How to use a laser cutter tutorial

• As a consequence of the previous session, several people have commented that the beam propagation time that I was assuming to be result of a cascade of physical atomic interactions, may well be mainly due to the response capability of the HV power supply.
• To investigate this in more detail I prepare a special set of pixel patterns at a fairly coarse 100 pixels per inch ( easy to calculate with)
• I set the patterns at various grey levels to investigate the dithering effect when we convert to black dots
• I am lucky enough to have two completely different machines. I ran these patterns at various speeds on both machines to look at the performance of two power supplies. The physics inside the tubes will be the same so any differences in results must be the power supply performance.
• Examining the results under a microscope for both machines at extremes of power and speed, shows the Lightblade producing much crisper “dots”
• It looks as though the decay time is the more significant factor

• Continuing with my investigation into bitmap engraving I realize that I am just scratching the surface of a very complex subject
• I can produce reasonable pictures onto most materials using luck and experience. However, I now realize that I do not understand exactly what is going on
• I have created a list of all the factors that I think are important to mastering that understanding A quick preview of this list…the next few sessions?
• In this session we will start by trying to find out the difference between theoretical and practical lens spot size.
• My vital USB microscope and measuring graticule.
• Using white card, and a fixed power pulse I try to establish the smallest practical spot burn for a 1.5” and a 2” focal length lens.
• Increasing power increases the spot size
• After tests on a static dot, I design a special pixel pattern (100ppi) so that I can run dynamic tests at various speeds and powers.
• All results show “sausages” rather than dots
• I plot the results graphically for each focal length to display how “sausage” width and length varies with speed and power.
• How to use a laser cutter tutorial

• Does RDWorks do something to my binary picture when it imports it
• Yes, it adds some grey pixels and can mess with the resolution if you resize the picture.
• I have designed some special 100ppi special binary bitmaps
• When I import my pure patterns I find the same grey additions
• Carry out physical tests on clear acrylic at different speeds and powers to check what effect that grey is having on dot depth/power
• First amazing observation is that black is NOT a solid line but still a series of dpi dots
• There are no noticeable differences between dot depths despite the grey colour of some key pixels. So, we conclude that RDWorks is NOT messing with the binary data
• Summary of photographic results
• Still more questions to be answered. The offset between left and right scan lines gets bigger as speed increases. Is it the line start or the line end that is being delayed?
• Back to my pixel pattern to add a shallow cut reference frame round the pattern to act as a static reference to check what is moving as speed increases.
• It shows that it is the START of each line that is being delayed
• Experimenting with the USER Backlash setting caused a one-sided offset.
• The lines can be aligned on one end but not the other….a strange problem. Yet to be investigated further.

• Look at the special 1.5” fl lens I use on the Lightblade machine so that I can do comparative testing with a more responsive power supply
• My incorrect understanding and misuse of backlash User settings
• Using the REVERSE INTERVAL to correct scan offsets.
• Summary of all my dotting discoveries
• How to use a laser cutter tutorial

• Review of REVERSE INTERVAL settings on my China Blue machine
• Using “beer mat” card as my drawing medium
• Turn air assist OFF for engraving
• Changing the picture resolution in RDWorks
• Using the RDWorks dithering system to get to a DOT picture.
• Head position set to bottom corner to give scanning UP the page
• Setting the parameters
• Slow mo video to see if beam switches off between dots
• 150 dpi picture turns out remarkably well
• Double resolution to 300dpi but leave interval at 150 dpi
• Under the microscope the 150dpi dots and lines look nice and even in tone and good dot definition
• 300dpi ….not good…..too light.
• Analysis of different parameters and results
• The stupidity of trying to get 600dpi to work
• Testing high resolution asymmetric engraving 600dpi (x),150dpi interval (y) and 400mm/s with 40%power
• The pseudo grey scale effect that this produces
• Slow mo of the pseudo gey scale beam intensity
• Microscope view of the reason why we call it pseudo grey scale It’s TIME regulated greyscale not variable power greyscale
• Reminder of what the pink beam means
• Summary
• How to use a laser cutter tutorial

• The difference between bitmap and vector images
• Is the dotting information going to be helpful for 3D engraving?
• Rapid control of machine power is the key to 3D engraving
• The problem of the tube’s non-linear power characteristic
• How I have linearized the output
• Generating a grey scale test pattern.
• Setting parameters for scanning greyscale
• Running the same greyscale tests on both my machines.
• Examination of the results
• Analysis of the results.
• Are my peaks due to pixel power?
• Not all the peaks match my pixel pattern. An interesting question, why not?
• Increase to 6 times the resolution but there is an exact replication of the pattern Why?
• Rapidly beginning to conclude there is no relationship between pixels and the peak pattern
• The same questions and patterns exist with Lightblade patterns.
• Are my peaks due to stepper motor ? Not sure at present.
• Trying to guess how 3D is working via RDWorks. Is there a software algorithm that is averaging pixels?

• Choosing a linear range from the tube power graph so that no correction will be required.
• First test results seem to be showing too much power for the test speed.
• Reduced power seems to be better. Still, lots of pixel peaks.
• What does defocussing do? It gets rid of the serious peaks
• Jump onto engraving a real 3D image
• 3 passes and the results are very good.
• Change the resolution from 300 or more down to 100dpi and changing to ash wood
• Not a good result. Lots of burning
• Try without air assist. Oops lots of flame
• Can a fast cut clean it up? Not really. But it does change the background texture from lines to pimples.
• Detail resolution does not appear to be affected by setting the beam out of focus
• The suppressed power range seems to work well.

• I plan to use my little USB Pico Scope to see what signals are going into the HV power supply to control the tube.
• The beam ON/OFF signal is NOT switching pixel pulses on solid black engraving
• Look at beam switching AND the PWM signal that controls the current flow through the tube
• No pulses there either!!!!!!
• OK ……I must admit that all my thoughts about pixel pulses were completely WRONG
• Let’s check the 3D engraving swatch
• Wrong again. No pixel pulses.
• Import the checkerboard pattern into RDWorks and it shows as a mid-grey rectangle
• The scope pictures show discrete switching for isolated pixels
• A solid black pattern shows no sign of individual pixels .
• A dithered picture reveals variable ON/OFF beam switching but constant power
• 3D engraving signals show beam ON/OFF at beginning/end of scan and varying current throughout the scan caused by the grey scale level.
• Are we looking at stepper motor pulses causing my phantom pixels?
• That’s another whole investigation
• How to use a laser cutter tutorial

• The controller talks to the tube via the HV power supply via just TWO wires.
• One to switch the beam ON/OFF and one to set/control the current level.
• Individual isolated dots can be seen to switch the tube on and off
• Greyscale can be seen to change the current level according to the grey scale pixel shade
• What is FACULA?
• Design a simple grey scale test pattern and demonstrate it working the red and blue wires.
• Set up the machine to SPECIAL mode and 50%
• The red signal shows 20khz PWM signal at a constant grey scale current value and the blue signal (which would normally be ON during burning) is switching on and off at 10khz with 50% on and 50% off
• Reset FACULA to 75%
• Now the 10khz blue signal is 25% off and 75% on
• Reset FACULA to 99%
• Yes, that proves that all FACULA setting is A) switching the normally constant beam on and off automatically at a rate of 10khz. B) Adjusting the % setting changes the ratio of beam on and off time
• How does this change the beam size?
• Explanation of how beam exposure TIME is important for determining beam size.
• Can I replicate this effect with DOT mode?
• No. Although DOT mode promises dots at 1khz max, I was only able to achieve 750 hz at best.

• This question arose from a dithered file that someone sent me which had also been set for greyscale engraving. Was this someone just ticking all the boxes in ignorance or was there a secret buried here worth investigating?
• Now that I can use the oscilloscope to view current levels and beam switching we should be able to answer the question
• Import a colour picture to RDWorks and it automatically turn it to greyscale
• Dither the picture and close examination displays a picture with 6 grey levels instead of the expected 2 (black and white)
• We are only running the program as a test, we are not actually engraving a picture
• First test is normal dithered settings with max min power set the same.
• As expected, current was constant, and beam on/off was switching to match pixels and pixel groups.
• Reset power to 10% and 50% and set output direct to generate greyscale engraving.
• Do we see grey levels in our dithered image? NO!!
• So, it is confirmed that RDWorks is telling lies when it adds greys to our displayed dithered picture.
• Starting again with the same picture but this time we dither with GREYSCALE rather than DOTS
• Now we can see current changes for every pixel, and we have therefore created a greyscale image for engraving
• Examination of the trace for a speed of 50mm/sec and 120ppi clearly shows the machine is cleanly changing current for EVERY pixel.
• What happens if we run at 200mm/sec (4 times speed) can the machine still respond?
• Yes. We can see 1ms per pixel for small grey scale steps
• How about 400mm/sec
• Seems too fast, so check at300mm/sec
• Pixels just about discernible so we should not attempt greyscale at pixel times of less than 0.75ms
• All looks good at the electronic signal level, but I can only guess what the output power from the tube will look like after those signals have passed through the power supply and the gas physics of the tube
• After making some assumptions I do a real-world test of a greyscale dithered picture
• Conclusion. Rubbish ….stick to dot dithering for photo engraving
• How to use a laser cutter tutorial

• This formulaic approach has been designed and tested for organic materials such as wood, card, leather, and low formaldehyde content MDF.. It may work on mineral materials but that has yet to be researched.
• Choose the shortest focal length lens that your machine will take.
• Choose your material
• Important to understand the principles of binary dot engraving.
• The ratio between the dots and the background is paramount for a crisp picture
• Organic materials have a scorching colour range whereas something like clear acrylic does not. It’s a binary material that engraves white. This is the opposite of burning and hence when working with acrylic you must make the image a negative before engraving .
• Set your lens to the ideal focal position
• I have designed a test pattern to help determine the smallest spot size.
• I use step gauges to set/test the best focus point.
• Use a magnifier to view your results
• How to decide on the dot size by using the test pattern.
• Converting that dot size back to a Pixels Per Inch resolution.
• Modifying the brightness and contrast of a picture
• Resizing and setting the resolution of the picture and dot dither.
• Setting the interval is very important
• CALCULATE the speed ……DONT guess it
• Reset the head position in the config system settings
• Explanation of what happens if you ignore the resolution rules.
• Testing to find the best power setting
• Cross flow of air and almost zero air assist is essential
• Examination of a very good end result.
• A look at another picture done on the Lightblade machine. One picture done with a 2” lens and a second with a 1.5” lens. Big difference. 1.5” lens wins

• Can we remove the pulses from the air assist?
• Making a crude accumulator to act as a pulse killer
• Result seems successful but when compared to the undamped flow there appears little difference.
• Look at the circuit diagram for the proposed air assist controller and all the fittings required.
• A few sessions ago I demonstrated marking stainless steel WITHOUT molybdenum disulphide, so I make a label for my switch using the same technique.
• Installation of switch, pneumatic pipes, and wiring.
• Set the Vendor settings to enable the blowing solenoid
• Everything working correctly
• With the cycle paused. Set the restricted air flow
• Test the manual override.
• Oops I’ve damaged my lens by engraving the stainless label BE WARNED
• Test of system with a piece of engraving and cutting MDF
• How to use a laser cutter tutorial

117 Hunting those Phantom Pixels

Show More

• How did my pattern of pixels get generated if not by pixel generated power pulses?
• I do believe that the stepper motor is at the root of this issue but that the patterns are a symptom of another problem interacting with the stepper motor.
• Crude demonstration of stepper /servo motor principles
• The importance of removing elasticity from a drive system
• Detail look at the design of the head drive system
• Sponginess of the rubber timing belt teeth
• Accelerometer tests were unsuccessful at seeing resonance
• Can we see system resonance with my crude mechanical resonance detector?
• Back to using acrylic as a means of detecting beam velocity
• Carry out a series of reference tests before making changes to the machine.
• Different belt material. Polyurethane with stiffer teeth and steel cords.
• Repeat and analysis of the acrylic reference tests
• Repeat of waggle meter tests and trying to relate stepper sound to the acrylic head velocity patterns
• We have proved that a rubber belt was producing resonance and that the new belt has narrowed those velocity spikes to a much narrower range.

• The design work has already been done for the Lightblade machine but for various reasons has not been fitted
• This session is all about adapting that design to suit the China Blue machine
• Examining the various critical dimensions around the machine that will dictate the design.
• Lots of boring CAD video with explanation for various design decisions.
• Programming the parts for cutting
• After doing lots of acrylic cutting of the parts the cooling water temperature is around 41C. Not over worried about the high temperature because power has not decreased.
• Assembly and gluing the parts together.
• Using one of the new mirror mounts for mirror 1
• Adding a red dot pointer to the assembly (the mk7 design.)
• Assembly complete ready for installation and testing next session.
• How to use a laser cutter tutorial

• Removal of the old clamps and mirror 1
• Jiggle the new tube mount into place without disconnecting the tube
• Make some in situ adjustments to miss tubes and wires
• Put fixing slots into machine
• Use the red dot pointer to test the alignment of the new mount
• Demonstration of how simple it is to align the beam onto mirror 2
• In an attempt to get the beam aligned onto mirror 3 my frustration at the troublesome adjustment system on mirror 2 finally boils over
• I design and manufacture a new mirror mount for mirror 2
• Success at mirror 2 !!!!
• Remove the red dot pointer and set the beam properly with the scorch method
• Setting the X and Y axis is simple now
• The importance of setting mirror 3 onto it’s “sweet spot”
• Aligning the Z axis
• Summary of the project success.

• Demonstration of the light pipe principle using a laminar flow water stream to show the beam being turned away from a straight line exactly like fibre optics.
• Discussion of what might happen when we fire a 60-watt beam through a 1.5” FL lens at the surface of an 80mm block of acrylic
• Test 1, Wow. Explain what you are seeing. How is this possible?
• The same tests with a 2” FL lens
• Observation of the effect of lowering the focal point into the material
• Discussion about the hot gas cloud the happens as the acrylic evaporates.
• The gas cloud effect that causes cut striations on the edge of acrylic.
• The difference between static and dynamic effect on acrylic.
• How to use a laser cutter tutorial

• Summary of what curtains are is there any way to get rid of them
• We concluded that the problem was a replication of the timing belt pattern
• Could it be that a lead screw fixes the problem? Too expensive and complex to implement
• How about a rack and pinion drive system? No, BUT this triggered a lightbulb moment!
• Let’s remove the timing belt
• Let’s reverse the timing belt ….Mad or not ????
• My temporary smooth roller bracket demonstrates the principle of using the belt almost like a rack and pinion system. and showed that the belt was a fraction too long
• The Mk2 design was made of steel and fixed a couple of problems identified with the temporary mk1 design.
• Because the belt has been reversed, the direction of motor drive has also reversed so the vendor properties must be changed to reverse the X motor.
• Various repeat tests of the curtain parameters demonstrates clearly that the CURTAINS have gone!!!
• A smug summary of my success

• This project will be two sessions. The first will be the design and manufacture of a motorless rotary engraver and the second will be testing it.
• I claim no credit for this design idea, I am just adapting the principle to a more suitable design for my machines.
• Yet another plug for the benefits of a solid steel table
• Step by step assembly and gluing the parts together (with tips)
• Having assembled the frame we now discuss the design of the special wheels that I am making myself.
• When assembled the wheels did not live up to expectations so I had to design a mk2 design with 2 bearings to remove the wheel wobble.
• Discussion and step by step assembly of the mk2 wheels
• Assembly of the live end stops
• Demonstration of the simple principles of this design
• How to use a laser cutter tutorial

• Our first test will be to engrave a biscuit making pattern on a rolling pin
• Creating the pattern in RDWorks
• We need a flat surface in the machine to make the device work
• Adding the pusher bar to the machine
• Why elastic bands?
• Understand the max size for your graphic
• Setting up the rolling pin in the fixture
• Setting the origin point
• Success 1 with wood
• Engraving a glass bottle.
• Principle of etching glass
• Tricks for glass engraving that people advocate.
• I use molybdenum disulphide spray as a heat moderator
• Choose a suitable photo image and modify it to a sensible resolution based on our previous photo engraving experience
• Set the rotary device to suit the bottle
• Set the origin and define the cutting parameters.
• Check the result. Reasonable 1st attempt
• Modify picture resolution and parameters for a much-improved 2nd attempt
• Great first step to exploring glass engraving along with the perfect proving trial for the non-motorized rotary device

• A simple test pattern
• Using acrylic to burn an unfocused beam track at different speeds
• Analysis of the results
• Examine the beam energy distribution.
• The effect of the beam depends on the absorption and damage threshold of the material that it is fired at
• Explanation of why line width changes as speed increases
• What happens to the beam profile after it has been focused?
• Can this be used to decrease dot size and improve photo engraving resolutions for “harder” materials?
• We shall find out as we explore different materials in future sessions.
• How to use a laser cutter tutorial

• I know that the Lightblade machine has a cut squareness issue when working with thicker materials such as 8,10 and 12mm materials
• Demonstration and analysis of the problem
• Rotating the air assist /lens through 90-degree increments shows that the problem remains the in same direction regardless of the lens position
• Using a red dot laser pointer to show how the focused beam is VERY sensitive to beam alignment
• What is the sweet spot?
• Having accurately measured the Lightblade head and drawn a detailed CAD diagram of where the lens axis intersects the mirror, I know that the centre of the inlet port is the correct aiming point to hit the sweet spot.
• A scorch test on the inlet port shows the beam to be about 1.5mm high.
• The only way to lower the beam by 1.5mm is to drop the tube down by that amount. Possible, yes, but that runs the risk of upsetting the whole beam alignment.
• With no vertical adjustment on this head mount I decide to shim the head UP by 1,5mm to catch the beam at the correct position.
• Using a centring jig, I establish that the Z beam alignment is now central and axially true
• Back to cutting a test square and evaluating the result
• Stressing the importance of understanding if the head design sets the sweet spot at the centre of your inlet port. DO NOT ASSUME THE DESIGN IS CORRECT.