RDWorks Learning Lab 108 Going Dotty Pt3

If you own a Chinese laser cutter this little series of videos about me learning how to use the free software provided, may solve the problem of trying to learn from a virtually unreadable manual.
I am nothing to do with RD Works, I am not an instructor and I am no expert. This series will document the essential bits of many hours of trial and error

[markramius]

Rapidly becoming a legend in the laser community, Russ. Awesome videos, helping hundreds of shed warriors the world over. Thanks mate!

[howdydoodyb]

I am really enjoying your hands-on "lectures". Some people are born teachers. It appears that you are one. Thanks!

[RickEverett]

Still....Great videos, and information, Russ.
Wish I could post pics on YT.
Did some engraving on pine, that turned out quite well, But, I'm no critic.
Brilliant work, please continue.

Rick

[Artiken]

I noticed in your variable length example, that the start power ramps up quickly, and is consistent on each line segment. The depth of the cut gets deeper, then slowly falls off. My thought and this would explain some of the difference between the acrylic and paper. The heated plastic causes, the same laser power, to tunnel deeper. As example, If you were able to set the initial power to 30 at the start, then drop to 20 to finish the rest of the pixels in the line, then off. The off time would be closer to the ramp up time of the ON. Just a guess. I have no idea how to accomplish this in the software.

Your video's are VERY informative. Some of the best I have seen online. Thank you.

[fredflintstone]

How to shorten your sausage.
I have two machines one china blue and a K40 the K40 was my first machine and I soon found out the controller in that was not up to the task I had in mind but more on that later.
I believe I can answer the issue of the tapered / variable hole depth you have shown in going dotty 3.
As a professional electron wrangler that is a recognizable wave form shown in the engraving depth.
There are two signals that go between the laser controller and the LASER supply.
One is the digital control this enables the supply to generate high voltage, So ON = lase Off no lase.
The other signal is Analog, this is a varying voltage that the laser power supply sets it's power to, in
essence a current control setting.

If you have a 80W laser supply with a 100% power setting of  40mA then putting a voltage of 2.5V on the
analog input pin will set the power supply to 50 % power,   5 volts will set it to 40 mA.
Now here is the problem this analog voltage is being changed every pixel.

There are several ways to create an analog voltage from a number, remembering that each pixel is a grey value. The way used in the most common controllers is PWM as Resistor ladder  D to A's are more expensive and have different problems. 

So the computer in the controller sees a pixel value of 127 and sets a PWM of 50% (assuming an 8bit system they might be 10 bit) this PWM is smoothed buy A resistor / Capacitor network buffered and presented to the analog input of your power supply. This method of analog voltage generation creates ripples on the voltage this is one of the reasons the bottom of your sausages are not flat.
The other problem with this method of voltage generation is the time to rise and fall. Looking at the wave
form shown in your engraving matches the decay / discharge of a resistor/capacitor network.
My summation of the sausage shown with 3 distinct power levels, (#9 in your diagram) indicates the PWM Pulses pumping up the
voltage every pixel then the PWM is discontinued, the capacitor discharges as seen by the shape of the tapering of on the left side(your pink line).
The controller I put in my K40 is an early clone of the rudia system and the software that controls it looks like rd works but is not. It
is some 6 years older than what we have now. In the K 40 I never connected the analog input from the controller, I cannot remember the
reason now, the power is controlled by a Pot on the front panel. I set the power by pressing the test button and turning the pot while
looking at the mA meter. In the software I have my power set to 100%

but I doubt it would make any difference If I changed it as it would only effect the disconnected
analog output. Using this machine I can engrave PCB traces as fine as 5thou (.127). when I get a chance I will add a switch and pot to my china blue as I cannot get the same results as I do with the K40. As a side note I engrave at 80mm a Minuit and only in one direction.
Having looked at the input circuit of a laser supply and found the first place the digital signal goes is a opto coupler I realized the
laser trigger time will be quite slow, opto couplers  are slow and they use the most “cost effective” ones around. Opto's also can
stretch signals as they can take more time to switching off than on.
They may compensate for this in the controller but I doubt it.
I will leave you with this brain pain for now. keep up the questions as we are all learning. 

Pr

[DoRC]

im really interested to see how these tests turn out on a material where the burn doesnt "spread" due to melting etc. im not sure what the best material for that would be. ideally it would be metal i suppose. at any rate seeing different materials is going to be very interesting!

[MikeAudleman]

I wonder. Those shark fins look remarkably like what I saw when I was doing my penetration tests and observed during the cuts. What you refer to as beam drag.

Question. Is it actually stacking the power or is it something else? Is it possible that the "stacking" as you call it is a product of material removal. It takes time for the material to vaporize. Could it be related to gas buildup inside the cavity of the pixel.

It would be curious to see an oscilloscope trace on the PSU output in comparison to the pattern in the material. That would determine if its the PSU/Tube capacitance in play creating the sawtooth pattern or something in the nature of acrylic and the vaporization process during the forming of the dots.

Another question to ask, is if you feed a true square wave to the laser tube, do you get a square wave of IR coming out? Ignoring delay, does the tube instantly begin emitting IR but have a gradual falloff after fall of the input power is switched to 0. And if it does, what is the falloff time of the tube's output power? I would hazard to guess that there is some sort of maximum switching frequency that these tubes are capable of. And the PSU, it must also have some sort of switching frequency limit as well. I doubt any of us are going to have the measurement tools to find that answer unfortunately.

Mine were produced by extremely slow (1mm/s) speeds. Clearly NOT related to PSU response time at that slow speed. So they have to be produced as a factor of the vaporization of the material itself. I would think anyway.

[RichardRocket]

You are the best of the best my friend...

[denniskramer]

Hi Ross - spot on (No Pun Intended)

There is a recent US Patent application from the folks at Rice University using a laser to create Graphene from controlled laser scribing. Their scanning technique found that polyimide (Kapton tape) worked best and that other materials PET, work at all. I suspect as you have shown - some materials 'char' and others vaporize. Those that char can create various forms of carbon - graphite, graphene oxide, graphene,
What surprised me was their example was achieved with much shorter pulse on times and smaller beam widths. i.e. 14usec pulse width and 120 micro meter beam size, and scan rates of .7 to 23.1 in/sec. They achieved there results from their machine running at a calculated power of 2.4 to 5.4 W. on a Universal X-660 machine.

1. If we're using card stock say for photos then the much lower 'burning' temp of the card stock vs. Kapton tape also contributes to the difficultly of a 'controlled burn'.

2. I think you are correct that one of (if not the main) cause of the rise and fall times of the pulses due to the power supply is the cause. The current will rise in the tube limited by the impedance (resistance and Inductance) of the power supply and its wiring. Also the switching elements (my guess high voltage mosFETs) also limits switching times due the charge (capacitance) within the semiconductor devices. Pulse trails can be cleaned up by providing a return path for the current (stored in the circuit inductance) when the mosFETs turn off. Choosing devices appropriately and PC board layouts can minimize turn on delays due to capacitance. Who knows what's going on inside our China Blue units.

3. What about any hysteresis affects within the on/off of laser tube. i.e. turns on a higher voltage than it 'turns off' as the power pulse decays?

So short of some serious redesign of the power and switching circuits - we're likely limited as you have shown to 100 ppi and 50mm/sec. (Any thoughts / ' experience with the smaller desktop machines - smaller spot size, better power control at the low end?)

Cheers,
Dennis

[bialyk]

Very interesting and quite close to my observations (on slate) - ref. to resolution. I suppose this "shark fins" are related to energy or in capacitors in power supply - or (less possibly) to N atoms energy in the gas in tube. You can as well change one more value during tests: frequency in Ruida software. It has influence on fast material cutting (tested ! in practice on very light wood cut fast). By the way - every time I see close-ups of cutting acrylic it remind me snowy, cold day in the winter and hmmm... pissing on it. Stream of power and reaction of snow. Act very similar. Energy cut material directly, but energy spread and cut goes wider when beam slow down.

[baraltayli]

Hi Russ,

Thanks for recording these videos and help us. I learned a lot from you.
I am just wondering is there any performance differences between lenses in terms of their material. For example If we analysed ZnSn vs GaAs lenses which one is better to use on CO2 laser?

Best Regards,
Baris.

[polakis]

Cut the glove label Russ😜 Nice video :)

[osamahamdy]

I need some help with Rd vision calibration