Ideally a hat should match the customers’ head size and shape. Obviously you can use a sewing tape measure to get the circumference of the head, but seeing that we are all unique flowers, we have been blessed with varied head shapes. Therefore, hat conformers are a thing.

They all work on a similar principle, multiple components that radially adjust to touch the head of the future hat bearer and lock into position. After that, the haberdasher can trace that outline onto whatever haberdashers trace their head outlines onto.

If you want to make your own, here are the files and instructions on how to assemble the thing:

https://www.etsy.com/listing/1841312248/hat-conformer-laser-cutting-files-dxf

Parts list:

• 1 x Laser cut – Base – 4mm plywood
• 23 x Laser cut – Element A – 4mm plywood
• 23 x Laser cut – Element B – 4mm plywood
• 23 x Round headed M6x20mm bolts
• 23x M6 wing nuts
• 23 x M6 large washers
• 23 x 5mm shelf pins/pegs

Assembly

1. Use wood glue to assemble Element A to Element B
2. Insert 1 shelf peg into base
3. Place assembled Elements into base and insert bolt from below
4. Place washer on top of Elements and Wing nut on top
5. Repeat 22 more times
6. An elastic band should be used around the vertical elements to make conforming easier.

On an unrelated note, have you ever seen an actual banana hammock?

The post What shape is your head? Laser cut hat conformer. appeared first on G00.

As a basis I used some kind of surgical clamp bought at the flea market a long time ago. I kept it around just waiting for an opportunity. Well, here it is. There was also a bit of 10mm flat stock in the workshop and used that to make the actual grabbing part.

After finding an appropriate diameter steel disc, I bent the flat bar around it, then used the wire bending jig I built a while back to bend back the ends of the circular sections.

Then I cut the end off of the surgical thing and drilled and tapped two holes for each tong. Of course I broke a tap in the process so had to drill another hole. So much for symmetry.

After a couple of matte black high temp spray paint, the tongs are done.

Here it is being tested out. Works great, and the built-in spring prevents you from dropping the flask.

The post Flask tongs appeared first on G00.

The next step we want to take is casting 3D printed rings. This would allow us to transition a bit more smoothly from silver to gold and we could make quite complicated designs with less effort.

One important note is that we wanted to keep the monetary investment as low as possible, so we took the cheap route. What I mean by this is that we could have just bought a vacuum casting machine and have everything already set up, but the cost is to say the least, prohibitive.

What follows is our first foray into the investment casting process. Certainly not a guide or tutorial but more of a technical diary for our future trials. Also note that I may have mixed up the terminology.

Act 1. The model

There are different ways of creating a model for a casting. One of the ways is by sculpting wax, for this you need a certain amount of skill which we certainly don’t have. Therefore, we went with 3d printing.

We bought an Elegoo Mars lcd resin printer and first printed some test rings in regular grey resin. After we worked out the bed adhesion issues, we got some exceptional prints. The adhesion issue was fixed after finding out that the more vertical the adhesion layers (the raft) the better the print sticks. The disadvantage is that they are a lot harder to unstick from the build plate after the print is done.

Here are some of the first prints.

Once we got that down, and once Victoria got familiar with blender, we took the next step and got some castable resin.
We got some Photocentric Daylight castable resin which was a lot cheaper than any other alternative, and got to it. I was really nervous about what settings to use with the castable resin but got it right on the first try. (80s for the adhesion layers and 10s for the rest).

Act 2. Putting it all together

After curing the prints in the UV lamp, we were ready to cast! Actually not quite. We still needed some wax sprues. Instead of buying wax in the proper diameter to fit the rubber cap of the flask, I just printed some sprues, then made a silicone mold and poured the wax in the mold.

The wax is actually used dental wax which we decanted.

Another thing which we used from the dental industry was the oven. It’s quite small, barely large enough for the flask, but we got it at a good price and it goes up to 1050 deg. C. The flask and investment powder were bought from a website that sells tools for jewelers.

Having everything we needed, we stuck the printed rings to the sprew and poured the investment.

Act 3. The burnout

Burnout is a terrible thing, but not in this case. With the flask in the oven, we tried to keep to the resin manufacturers’ burnout temperature chart, which said that first you need a 6 hour 150 C stage, then an 8 hour 780 C actual burnout stage, after which you cool down to 600C. Then you have a 2 hour window to pour the metal.

The truth is, by the time we reached the 14 hour mark, it was already 1 o’clock in the morning, so we just turned of the oven and let the mould cool down overnight. Our research revealed that some people allow their flasks to cool to room temperature and cast then, so we thought it would be okay.

Because of the small gap between the flask and the top of the oven, the airflow is very restricted. This airflow is needed for the gases resulting from the burnout of the wax and print to escape. Some ovens have air pumps to aid airflow.

Another issue is the cracking of the investment. There are multiple possible causes, but I’m willing to bet that turning off the oven directly from 800C without slowly turning the temp down, was the cuplrit. Nevertheless we pushed forward and cast the silver anyway to complete the experiment. Mainly to find out whether all of the model burned out cleanly.

Act 4. Casting

As we usually cast ingots for wire and sheet, we have a Sievert MAPP gas burner and a crucible which we used for the pour.

And here is the result.

There are a lot of things wrong here, but it’s a start.

• Incompletely filled mould. This was probably caused by the room temperature pour and the lack of more inner sprues to guide the silver up to the top of the rings.
• Rough surface. I’m assuming this is also caused by the cold pour.
• Air bubbles. One of the rings has sharp edges and they are not filled completely. This is either because of the orientation of the model, the cold flask or the lack of a vacuum pull while pouring the silver.

We’ll try once more with a 600C flask temp when casting and see what changes. Fingers crossed!

The post First try with investment casting appeared first on G00.

The post Welding mask appeared first on G00.

So, with Mach3 out of the picture, I decided to dual-boot LinuxCNC. The problem with this was, mainly, that I also have a laser cutter that comes with RDWorks which only works on Windows. This already meant a loss of productivity because I couldn’t use both machines at the same time. Nevertheless I got LinuxCNC to work, but then after having difficulties getting used to it,  and realising that I wouldn’t be able to use Solidworks or Fusion360 to CAD/CAM the parts, I finally decided to convert the machine to USB.

In order to do that, I whipped up a quick USB to parallel port adapter using an arduino nano on a perfboard and installed GRBL on it. You can get it here. The schematic of the board doesn’t exist, because basically it’s just GRBL pinout -> Driver board pinout. All wires.

I’ve used GRBL to drive DIY CNC’s before, and it was nice, but it had one big shortcoming. The feed override during machining wasn’t happening in real-time. So you couldn’t adjust the feed while milling difficult materials except after the current running command was finished. The latest version (1.1f) has this capability and from my point of view, GRBL is now perfect.

You can see it hanging on the left of the picture off the parallel port of the driver board.

It’s still missing limit switch, probing and spindle PWM connections, but it’s a start. By the way, the original spindle controller PCB has PWM input capability so it’s just a question of adding a connector from the arduino adapter board to the spindle controller.

Having converted the machine to GRBL meant that I could now use bCNC to drive it. This is an excellent piece of software, also free, also open source, written in Python. It has a bunch of features including generatic g-code from dxf files, making text engravings, adding tabs to the g-code to keep the cut part from flying away and breaking your mill. You can get it here. The user interface is very friendly and I cannot recommend it enough! Here is a screenshot from the bCNC github:

Here’s a test I was able to make with the Fusion360 – bCNC workflow:

(about 50mm/min, 10.000rpm with a 2mm, 3 flute endmill with WD40 as coolant)

Stay tuned for more CNC related posts!

The post GRBL+bCNC=Love appeared first on G00.

The post Electronics components organized. appeared first on G00.

Design-wise the course I took was to use a lamp boom/arm and mount the fan on that, so it could be positioned as close to, or over the soldering job.

Half an hour of solidworks later, I had 3 laser cut parts. Two grills and a spacer consisting of 5 and 3mm acrylic to get to and 8mm space to accomodate 2 sheets of filter material. With the filter I went the only way I knew which was using an activated charcoal foam sheet which is sold for exhaust hoods. So I cut it out in the shape of the spacers.

Then I just dismantled the cheapest boom-arm lamp I could find and mounted the fan onto that. Of course it took me three tries to get the holes right. Tired, I guess.

All assembled, and waiting for a speed controller:

On to the speed controller. I briefly thought about using an Arduino to provide the PWM for the 4 wire fan I used, but then I realized I hadn’t used a really old friend in a long time. My good old buddy, the 555 timer. I found a schematic online to provide the high frequency PWM for 4 fire fans and built the schematic on a small perfboard after trying it out on the breadboard first. This way, I didn’t need any voltage regulators and I saved loads of space.

I’ve fallen in love with this lamp arm. It’s so cheap (around 5 euro) and versatile! I also built a microphone adapter for it, for a friend of mine.

The post Fume extractor on a boom-arm appeared first on G00.

Laser cut with 0.1mm outside kerf compensation to get that “hammer” fit.

Added some cheap window gaskets:

Then I just cut two square shapes to hold the PVC sheets. Turns out this was a bad idea because of the way that plywood bends under heat.

The way I tried to heat up the plastic was with a cheap electric radiator and a hot air blower. The radiator didn’t have enough power to get the plastic soft enough and the blower wasn’t uniform enough and kept either not softening the plastic at all, or just completely melting it.

Anyway, this was my best attempt:

Gotta go buy me some of that fancy HIPS I guess. Maybe a more powerful heater too.

The post Vacuum forming. How hard can it be? appeared first on G00.

Doesn’t look like much. Covered in paint, absolutely filled with sanding dust and and it had a split belt drive pulley, but i managed to clean it up a bit.

Finally a quicker way to get those aluminium extrusions square.

The post Bought a belt sander. Cleaned it up appeared first on G00.

Hardware

The belt was scavenged from something a while ago and the pulleys are laser cut acrylic. They both have the same number of teeth because the precision of the stepper motor and microstepping would be enough for the small diameters I would be working with.

A quick calculation shows:

A resolution of 1.8deg/step results in 200 steps/revolution. For a 10mm diameter stock, this equates to 31.4mm/200 = 0.157mm/step of surface displacement. With 1/16 microstepping factored in, you get 0.0098mm/step. The resolution I use to engrave is 0.06mm. So more than enough.

Having dismantled an old cordless drill a couple of months ago, I had the chuck and the original shaft of the gearbox. Luckily it had three prongs sticking out, which I used to mount the pulley on. Upon realizing I had no bearings of this size in stock, I went to the autoparts store next to my workshop (it was a late Saturday). Luckily, alternator bearings matched the shaft and they were cheap, too.

After using some DB9 connectors to easily exchange the motors for the Y axis, I was good to go.

Here’s what it looks like (I’ll upload some non-potato pictures when I get around to it):

Edit: Here are some more pictures:

It may not have been such a good idea to use plywood for this build but it is stiff enough for now.

IMPORTANT: Turn your laser cutter OFF completely when connecting/disconnecting stepper motors. Not doing this may damage your stepper drivers.

Software

On the software side of things, the only thing I had to do was enable “rotate engrave”. In this screenshot, the option is inactive. I’m not sure why, because on the workshop PC it’s active. In the workshop I have the original “Laser CNC” software that I received on the CD with the machine; on this laptop I’ve downloaded RDWorks 8.01.18 and the UI is a bit different, and the machine is not connected.

“Circle pulse” here, a bit obviously is the number of steps for a complete rotation.

Also, the end-stop for the axis which is replaced (for me it’s the Y axis) needs to be manually triggered twice during machine initialization.

Action!

Here’s a GIF of it working on it’s first project (write-up coming soon!):

Thanks for reading!

The post Rotary axis for the laser cutter appeared first on G00.