Laser troubles

March 11th

I feel as it is Week 7, a little housekeeping is in order.

Here’s a list of all I have done and learned throughout the project so far:

Week 0

Applied to the Scottish Design Exchange placement, and proposed laser cut plywood puzzles, jewelry as a product

Week 1

Idea of miniature self-sustaining terrariums
Proposed micro-terrariums using moss, stones, branches, tiny crystals, in small jars
Intent to use 3D printer and laser cutter at home

Week 2

Collected moss and rocks
Created first prototype of micro terrarium with moss in small jar – earring
Research on preserved moss
Goal to design something beautiful, unique and useful, environmentally friendly USP, awareness to global warming, species extinctions, and excessive pollution
Terrarium idea confirmed – LEDs, open enclosure
Set up laser cutter – engraving test
First logo sketches
Found out about the San Rafael waterfall in Ecuador
Sculpture shape sketches
Topographic laser cut sculpture idea
Idea to find plywood offcuts from a local source to minimize emissions and waste
Laser cutting test

Week 3

Fibre optic LED light, laser-cut geometric open terrarium – no extra fuctions
Living hinge dome idea
Bought electrical components, USB cable to power
CD “rainbow” idea
Shape sketches
First CAD design (Fusion360)
First cardboard 1:4 scale model
First soldering test
First Adobe Illustrator .svg layers

Week 4

Electrical Health and Safety reminder
Wires for dome
Preserved moss
Second CAD using .svg layers – decided on 2mm plywood
Research on electrical circuits
Vinegar-steel wool wood staining research, test
Testing wire twisting for cage with 1mm wire
Wood staining test

Week 5

Wood stating after a day: green tea + normal white vinegar + steel wool is = grey-blue
The collected stones and branches got moldy
Shop shelf display design testing
Idea for large tropical leaf on the shelf display
Wood glue testing with plywood, moss, coarse sand
Electrical circuit questions, research, calculation, parts purchased
CAD circuitry design
CAD assembly gif
Shaper Origin Fusion360 Add-on
CAD layers to .svgs to Lightburn
Miro and ChatGPT to find a brand name
Walk to find moss and rocks – cleaning, preserving
Shop shelf layout mockup – leaves DIY
Twisted wire cage test

Week 6

Bio writing
Chrysalis brand name
Testing shelf layout in store – decided on stones and dark leaf
Chicken wire dome suggested
Bought 3mm aluminium wire
Soldering test
TPU 3D printing test
Vertical tab designed to keep button secure
Wire design in CAD
Quick Photoshop mockups of chicken wire dome
Decided on no chicken wire – 3mm aluminium cage
Strain relief CAD design, TPU printing test, redesign, test
Fixed CAD assembly gif
Soldering test – successful
Multiple strain relief prints – unsuccessful

That was a busy few weeks for sure.

March 12th

CHRYSALIS

Chrysalis (which covers and protects a developing butterfly) encapsulates the idea of emerging as a new, better version of ourselves.
My name is Krissy, I was born and lived my first 23 years in Hungary but moved to Scotland in 2016.
I started my product design studies in 2019, when I found my calling. I’m a third-year university student now; this is my first time selling something I created.
I am passionate about designing products that cause the least harm to the environment. Our planet is in a fragile state, and designers have a big responsibility to help it heal and thrive.
The area depicted is called the San Rafael waterfalls in the Ecuadorian Amazon rainforest – one of the most biologically diverse parts of the planet, and the “lungs of the Earth”.
The waterfalls, green hills and rocky faces are all contrasting and yet beautifully working together to form the perfect diorama.
Created out of laser-cut plywood, hand-picked and preserved moss, and natural coarse sand, with a gentle LED glow, it reminds us that we can do more for ourselves and the planet, and learn to be in harmony with Nature.
Moss positively affects cognitive function, stress levels, and emotional well-being, and it helps to draw moisture and airborne particles out of the air. Take a deep breath while admiring its beauty, and continue your day with renewed energy!
A portion of the proceedings will go to the WWF Amazon appeal because I believe every little bit helps to conserve our Earth. (https://support.wwf.org.uk/donate-amazon)

March 13th

Yet another day spent testing the laser cutter with Lightburn. After many videos watched, posts read and Facebook posts written to dedicated groups, I found out that I should use the Origin setting, never Home, I managed to frame the test grid I altered from a file offered on a YouTube video without the alarm going off.

Unfortunately, I could not find any information on 2mm bass (or balsa? The Amazon ad is unclear) plywood, so the test cut through the numbers which I intended to only engrave, and the first square started the actual flame alarm on the machine. I also somehow accidentally mirrored the image in Lightburn.

I am stressed out of my mind because the thought of starting a fire scares me so much. But plenty of people have managed to do this. I am sure it can be done safely.

I am now going to trust the official X-Tool material recommendations – although the thinnest they have is a 3mm plywood (cut at 100%, 360mm/min, 1 pass, and engrave at 80%, 3600mm/min), I can try to lower the power or elevate the speed, and maybe drink some Chamomile tea to calm my nerves.
Sometimes a cliff with nobody around for miles would really help to get my frustration out by screaming.

Two gentlemen on the Facebook groups answered my cry for help with recommendations of 80%-90% power, and 300mm/min as well as 800mm/min speed with a single pass. Thank you, kind people on the internet!

Here’s the next test I started – the intention was to score, not cut through which I did achieve, however, the “cut” is still much too strong – I suspect the lower time is the culprit, and possibly the focus.

March 14th (Happy Pi day!)

Today I learned how to take off the laser head and clean the lens. It wasn’t a necessary step, but I cannot figure out how to focus other than using that option, and I thought I might as well check the lens is still intact. I have some fancy replacement lenses from Hemsley, but I would have to take the previous one out, which is still fine, and that may void my warranty, and I might damage it. So I’d rather leave that until I absolutely must. I did use a cotton bud and some Isopropyl rubbing alcohol to clean it though.

I have great news and silly news!

The great news is that the cutting worked great. I had an initial test when I couldn’t see if it was doing anything, and I stopped it prematurely – but the 500 mm/min speed with 12% power worked wonderfully for etching.

I did look into how to cut text with a single line – turns out there are some fonts in SHX type (used by AutoCAD for CNC-ing – some kind person on the internet put together a folder of about 40 of them, and I chose the one called CIBT_, but I later found a program from 2001 called SHX Viewer 1.1 on a rather dodgy-looking website which allows me to preview them – looks like it’s only a 30-day trial), that, with a certain setting on Lightburn, I am able to use and not have to cut around the letters, which takes longer and makes thin lines looks burnt.

In the second test, I got to the first three lines of the test grid, actually cut out some squares, when (and here’s the silly part) the whole plywood plate flew up – it was either the air assist + fan combination, given how light the 2mm plywood is, or the fact that I changed the height of the laser head, which, due to the warping of the plywood may have touched the corner and moved it. It was quite scary so I immediately stopped it, but in hindsight, as nobody got hurt and nothing got ruined, it was rather funny.

According to my latest test (which I only stopped at the end of the first line of engraving because it seemed to be burning it too much), 600mm/min at 60% power is great for cutting, I will try to bump up the etching speed to 600mm/min, and keep the 12%, while for engraving 6000mm/min at 50% seems to do the trick.

As a side note, based on what I just learned through all this testing, if I were to create another testing file, I think I would make sure that I write the power and speed on each of the squares (or whatever shape I decide to use), because since they are small, it is difficult to tell which combination gave me which quality of cut, they get all mixed together when I take the piece off the cutting plate. And the other thing would be to leave a little more space between engraving, and to make the shapes slightly smaller, so it takes less time.

Now to cut the project files! This will of course require some redesign, because of the new settings and fonts I have found.

I should also think about what I will engrave/etch on the bottom of the product – which is why it is probably a good idea to test some fonts with my name or the brand name Chrysalis.

For numbers, I think the SHX font DINS is great, while for the brand name, CIBT_ or SCRIPTS8 (with an angle of 21 degrees).

The (hopefully) final Lightburn settings on the first cut file:

Upon cutting the first of 8 pages of the project layers, I have made the decision to up the cutting speed to 750mm/min at 60%.

Which did not work, as it didn’t fully cut through the material, but made the cut look an unappealing yellow.

So now I will try 800mm/min at 75%. The etching is fine, that just needs to visible enough. Also, what is with the framing showing me that the piece fits perfectly onto the 200×300 plywood, but when it cuts, it doesn’t?!

While researching this issue I watched a video where someone suggests to turn off the air assist option on Lightburn, as mine is not controlled by the machine, and it would just add extra GCode (which I assume means added time and energy).

To make sure it will cut exactly where I want, I will need to put the frame as the first layer in Lightburn, turn on the output and change the power to 0.1% (which turned out to be so low I couldn’t even see the laser), so that it doesn’t cut, but I can see where it will be.

Next try: setting the frame to 3000/2, clicking output on this but switching off the other two layers, positioning the laser by only firing at 2% to make sure everything fits on the plywood, switching off the frame layer, switching the etching and cut layer back on, changing the cut layer to 750/80 because it STILL didn’t cut through, and praying. I now positioned the laser head to the exact back left corner, instead of moving it slightly off, and made sure it all fit on the sheet, so now I shouldn’t have to move the honeycomb frame at all. I can simply put the plywood into the corner groove.

Nope, no luck. Still didn’t cut through. Also, I wanted to check how long it takes to cut one file, because for some reason Lightburn decides that the cutting project ended about a minute before the laser actually stops. Go figure.

Next try: 700/85. Please. 7th piece of plywood MUST be the charm, right? And I will try to re-use the ones that did not fully cut through. I promise. Maybe use a scalpel to go around it once more by hand.

Layer 8 – 19.05 pm – 19.10, so it takes about 5-6 minutes for this layer to cut.

What I do not understand is why it isn’t cutting through the curves.

I should have been smart and try to use a piece from earlier to do a test run with some small arched door shapes, those have sharp corners as well as a curve. It would have helped me to dial it in, without losing seven sheets of plywood. But it’s a learning curve, right? I was just so sure that I got it right. Each time. It is my first time using a diode laser cutter after all. I should cut myself some slack.

I did this by creating new layers with the different settings, and disabling the actual pieces.

Aha! Some actual, useful information finally. Looks like I will stay with 600/95, and have tiny tags still connecting the pieces to the frame, which actually isn’t that much of a problem – I would prefer to be able to take the whole sheet out and easily knock the bits out from the back, without much damage. For this project, total micro millimeter accuracy isn’t necessary.

What a fool I am to actually think I have tamed the beast. While the small “gravestone” shapes cut fine, the large pieces did not. This may be because of the slight bend in the plywood, or impurities in the material. Or any number of other things I am not aware of.

The magic numbers are 575/97 with 4000/2 test framing and 800/20 etching. Trust me, I’m an expert.

Alright, alright, maybe 20 is a bit of an overkill. 800/15 it is.

So in order to try to save material and time, I drew the archway test pieces on the same sheet as the pieces I intended to cut, in the empty spaces. Which was fine, but unfortunately when I put it back to do this, the framing was incorrect, and had to redo 6a and 7a.

When cutting Layer 1, I actually noticed the holes for the LEDs were missing. I must have accidentally deleted them, but it was a quick enough fix.

Unfortunately, I only noticed later that I copied the holes onto the wrong side.

When cutting it seemed that the laser cut through Layer 5 twice – I’ll need to make sure to turn the setting on in Lightburn which only cuts double lines once.

There is some strange white powder on the plywood that I attempted to quickly brush off before placing them on the honeycomb to cut, but sometimes it was excessive and I couldn’t see very well in the dim light.

I tried to research this but most results were pointing me to old houses with white mold on them. I seriously hope that is not what I am dealing with.

It looks like I somehow missed layer 4c on the sheets. But there is a 4c and 4b which are the same. Oops. Here are all the mistakes I have found:

-7 needs outlines for the layer on top
-2 needs holes for the wires (provided I stick with the 5 pieces of 3mm aluminium idea, which I am not certain of)
1 has the wholes in the wrong place
5a, 5b, 5c need outlines for the next layer

I am also going to go through all of the pieces and change the orientation of the numbers so that it faces the front. It will just help with the assembly time.
I noticed one of the tiny pieces, -6b was missing – it is one of 3 I intend to use to have something between the button and the USB cable – I found this when I took the honeycomb off the steel sheet.
I then proceeded to clean the steel sheet because it was nasty. I’m not sure what chemical processes happened while laser cutting, but all the soot and yellow discolouration was stuck to the sheet, and simply using a paper towel with 99.9% isopropyl alcohol was not helping matters much, which is why I had to use a wet wipe first – this made quick work of it. I wish I took a photo of this. Perhaps next time. I will need to remember to clean the laser lens.

And here are the fruits of all my labour, temporarily put in place. It’s nearly midnight, but we got there, baby!

I did have two pieces of Layer 15 in there that I only noticed later.

I will have to change the cutting time. And maybe up the power to 100% for good measure. Some curves are just uncomfortably not cut through and leave splinters when I try to separate them. Yikes.

550/100 it is. For real this time.

March 15th

In attempt to save myself the trouble of firing up the laser simply for two sheets pieces of re-cuts, I tried to cut the missing holes on Layer 1 and -2 by hand, using a scalpel and a small hand drill. This worked fine for Layer 1 as the holes were for the LEDs and were bigger. However, I broke Layer -2. So I have to fire up the laser anyway. I would like to glue the product together according to my CAD gif, and I want to make sure it has everything I intended to be there.

So that’s what I did. And then I broke the laser cut one, too. The curves are still not cutting through fully and it just just so happened to be the worst where the hole was closest to the edges. But it’s okay, because I’ll be using wood glue. I will probably redesign the piece for the next prototype.

I will change the cutting time to 525/100. Dang it. But I will only try this out when I next cut layers. I will first glue these ones together. I still need to find the best method to stain wood light blue, but for this prototype I will not bother with that.

While putting together the layers with wood glue (which is the most fun I’ve had on this project since I started it 7 weeks ago) and inserting the working soldered LEDS with the USB cable, I realized I forgot to put the strain relief on it before soldering it. So I had the bright idea to cut the strain relief and insert the cable that way. Which looks fine, although it doesn’t have the strength I designed it for. Maybe I should not have been too lazy to cut the USB and re-solder.

I found that I forgot to put layer -7b (the backing to the button) on the sheets, so I rectified these and quickly cut one out. I’m not certain simply gluing this in with wood glue is going to be strong enough to take the force of the button pushed.

After inserting the LEDs I tested them, as well as after adding a few more layers above and underneath.

Unfortunately, my initial thought that adding the fibre optic early on to help with alignment was correct, because I didn’t do this, the alignment was off and the small waterfall LED light is not visible, while one of the two holes for the fibre optics on the large waterfall does not go through fully.

The result is a very dim light on the large waterfall, and none on the small one.

When adding the fibre optics, I realized that this material due to its stiffness (despite the 1.5mm diameter) does not allow for the intricate curves I designed for the waterfalls. The small one is fine (although I need to be careful to make sure the layer on top is properly glued on), but the large one doesn’t follow the shape of the river.

Although the piece looks nice when all the plywood layers are glued together, it really comes alive with the coarse sand and moss. I used the wood glue for both (for the sand, I simply sprinkled it on the wet glue and allowed it to dry) which worked nicely. I did use up all of the type of preserved moss I prefer, which tells me it is not very sustainable.
I tried adding the 3mm wire cage I created for the previous quick card model, but upon talking to a friend who had a few suggestions, I removed it and made a decision to find a better solution. I have some ideas, like using the large sheets of dark brown Polybak I have, which is a very thin material, that I believe I could laser cut a front-open geometric frame dome.

I do not particularly like the framing edge on Layer 0 (where the stones and moss start) which is meant to be a continuation of the space for the holes where the wires sit, but it would probably take quite a while for me to redesign where the button would sit. Upon recommendation, it would make more sense to make the top layers bigger. But then I would have to re-align the layers on the sheets, and I would need to use more than 8.

Adding some more coarse sand to the front where the layers do not align (on purpose) might make it look more realistic.

I also had the idea to add very tiny animals such as a flying bird, but I feel the size would be unrealistic.

The etching on layer 4b is not necessary, and even the wood staining may not be. I had a bag of open kapok bedding (the softest thing I’ve ever touched) which I purchased years ago for my darling belated pet mice, but never gave it to them because I was told it was dangerous for them. I have been wanting to use it for this project in some way, because kapok trees actually grow in the Amazon. When the model was nearly finished, and I compared it to the photograph, I knew it would be perfect to imitate the water vapours at the base of the waterfall. I also tried it for the river itself, and I thought it gave a nice transparent elevation, but ultimately most people would probably say it looks “furry”. Which is not the intended effect of course.

My partner suggested adding a plaque to the front of the base and I thought it was a brilliant thing I didn’t think of because I’m too close to the project.

I used the edited Fusion360 model (which I cut the top layers to be level, and filled up the wire holes) and first added a slanted plaque with faux engraved screws, then a slanted one without these slightly lower, and finally one embedded into the front of the base with a curved outline.

The final plaque reads: San Rafael Falls, Ecuador – Made with love by Chrysalis

I’m thinking I would leave that last part off, or engrave it into the bottom piece. When someone looks at the product, they will probably want to see what it depicts, but not who made it. So I made a version with only the waterfall’s name, and a simple curved faux frame outline.