Wednesday, September 2, 2015

Sense 3D review

Ok I didn't build this one, I just bought it. It's a portable scanner, very easy to use, produced by Cubify. This is a link to their website.

If you want to buy a second hand like me make sure they give you the activation code as well, or you won't be able to use the software. I bought it for about 400$ (AUD) when new it was around 600$ (even cheaper now), including software activation (you have to download it though).
I tried scanning many different objects of different sizes, and some people, including myself. As you can see from the pictures, it gives its best with people and medium objects without straight surfaces. When used with small objects some details are missing.
The software interface is really easy to use and it shows the 3D model during the scan, so you can see the gaps and scan where needed. Sometimes it loses track, so you have to go back until it finds it. It usually happens when there is not enough light. I think it works better daytime, when the light is diffused and not coming from one source only. Light affects the quality of the scansion.
The software allows the user to modify the model, cut it, get rid of some parts or defects, improve the colors.
In the following pictures you can see the items I scanned, the Sense 3D software, and the final stl models (opened with STLview).  




These are some smaller objects


These are examples of people I scanned and printed. It's very realistic, detailed and impressive!




Conclusion


If you need a scanner for small object or precision applications, this is not what you are looking for, but if you want to have some fun scanning and printing your friends, this is the cheapest option on the market and it's really easy to use.




Friday, July 17, 2015


More about 3D printing


I think the best way to learn 3D printing is to make something useful, functional, something you have to design yourself to match with other components.
In my work it happens pretty often to break these small connection boxes at the back of the flood lights and we can't just buy the box to replace, we have to waste the whole light.
I decided then to print some of these connection boxes. I started measuring the existing boxes, thinking how to make them stronger and drawing it on Autocad.  

WARPING ABS

Obviously they have to be as heat resistant as possible, so I decided to print in ABS. The biggest problem I had with ABS is called warping: ABS shrinks when it gets colder, so if you have a large area, the layers above will shrink causing the corners of the base to curl up from the plate.
Some tricks can help with that:

- keep your bed hot, ideally around 110C. I used a IR thermometer to check the bed when hot and I found a couple of colder spots. I try not to put the edges of the base in these spots. To do that I offset the center of my print with the slicer software.

- clean the bed: if you print on glass or aluminium clean your plate with acetone before every print, ABS won't stick well on a dirty plate.

- I tried to print on paper tape and it sticks very well. It was even difficult to remove it after the print

- Print on the small side: warping is proportional to the area of the base so, when possible, make your base on the smallest side

- ABS juice: that's messy but effective. You will need to make this juice, melting some ABS in acetone.
And spread it on the bed:
There are some instructions about how to make it and how to use it here:

You can make denser ABS juice (ABS slurry) and use it to fill gaps and fix cracks as I did here:
You don't need to make a mess as I did, just fill the crack.
I also used ABS slurry to reinforce the base of this vase on the inside (I just poured inside and spreaded), but the acetone in the slurry melted the thin base and made it stick on the surface where it was standing (a paper towel).
When you use ABS slurry remember there is acetone in it and it will melt the weaker parts of your print, before it gets solid all together.

KISSLICER SETTINGS

I learned also how to change some settings on my favorite slicer to prevent warping:

- Infill width: if you make it smaller than your extrusion width (the diameter of your nozzle), it will leave a small gap between the extruded filaments (when printing with 100% infill). This will make the part weaker, but less prone to warping.

- Brim: if you write any number in this box, it will print a base around your printed part, to increase the grip on the bed. The number you write is how many millimeters this base will be larger than your part on each side. The brim height is how high this base will be. If you tick the "fillet box" it will start with the brim height attached to the part and it will gradually decrease to one layer to the outer loop of brim. I often use a brim diameter 10, height 5 with fillet, as in the picture:
In the same page you can add a skirt: it's a loop around your print, just to get the extrusion going before starting the print of the part.
Another option I change often is the number of loops (in the perimeter of each layer): if I'm printing an empty or partial empty part, I need 2 or 3 loops on the perimeter to make the part strong, but sometimes they don't stick on each other very well, causing the outer skin to peel off (on straight vertical surfaces), like in the picture:
When I want to print something strong I use 100% infill and only one loop on the perimeter. I might have to draw the solid exactly as I want it printed.

Another very effective way to reduce warping is to make sure the first layer is weel spreaded on the plate. To do that I set an offset of 0.1mm on the Z axis. This way the noozle touches the plate and spread the plastic on the plate. The first layer will look messy during the print, but it will be fine later, and the part will stick on the plate much stronger.

ACETONE SMOOTHING

ABS melts with acetone. The best way to use acetone to smooth the surfaces is in form of vapor. I pour small quantity of acetone in a glass jar or small container (make sure it's not affected by acetone). The ABS part shouldn't touch the acetone, so It must be standing on a platform or hung from above. But when you make your supports, consider that your part will get soft and could break around a hook or stick on a platform (it all happened to me). I like to smooth transparent ABS because it gets very shiny and you can see through it.

- Cold: to make sure the vapors are spread equally around the part, I cover the side inside the jar with toilet paper, which absorb acetone and make it work on the sides of the part. It takes about 30 minutes to 1 hour. In the picture, the big one is just printed and the small one is smoothed with cold acetone vapor: 
- Hot: the hot vapor rises all around the ABS part equally and usually 10 minutes are enough. Some weaker parts can melt completely and fall off. You have to keep an eye on it because one minute too many can melt too much. In the picture, the left one is just printed and the right one is smoothed with hot acetone vapor:
Either way you shouldn't touch the ABS part until it gets hard again. Also, be careful: acetone is flammable and dangerous.
Not only the part will look good, it will be also much stronger as the layers are melted together on the surface.
I tried this process on one of my connection boxes and I noticed that it shrunk a lot in the process, to the point that it doesn't match with the other components. I measured about 2% shrink on all dimensions.

RESIN SMOOTHING

Sometimes I print in PLA, because it's easy. PLA is not affected by acetone so there are different ways to smooth the surface. I bought a kit of XTC-3D from 3D Printing Studios in Fremantle. It's some kind of epoxy resin with longer cure time.

Here you can see a review of this product from the guy who sold it to me.

It's a very strong coating, slightly flexible and light. It doesn't affect any plastic, it just fills the gaps between the layers, making the surface smooth. It can be sanded and colored with acrylic. It has a good effect on layers 0.2 mm thick or lower, but if you print 0.3 mm and over, you might have to apply it twice.
In conclusion, it doesn't make the part stronger like acetone vapor, but it creates a very hard surface around it, and it doesn't have the problems of acetone smoothing (deformations, shrinking, hazards).
In this picture the large crystal is coated
with XTC-3D and the small one is smoothed
with acetone vapor. The look is very similar.
I applied a lot of XTC-3D on this vase but it
needs a second layer for perfect smoothing.

CONCLUSIONS

After all these problems and experiments I'm finally happy with my connection boxes, I just need to test them for heat resistance and insulation.


1 - Printed
2 - Assembled
3 - Installed
4 - Wired
5 - Tested



REFERENCES

Wednesday, March 11, 2015

3D printer!


I can finally write about my best achievement so far: I built a decent quality and large volume 3D printer. It's an open source design of a delta printer called "Kossel", delta means the three motors have a parallel movement.



The advantages of this kind of printer is that each motor has a similar load. I won't give you instructions about how to build it, there is enough material on the web already.
I will tell you where I bought the components, where to find the instructions, what problems I had and how I solved them. Handmade printers require a lot of calibrations that involve mechanical, electrical, electronic and software adjustment. So if you are not too confident in these fields think about buying one instead of building one. There are some decent quality 3D printers on the market under 1000$ (AUD). Sometimes I'm not patient so I skip some important steps and I end up burning something. This kind of project doesn't allow for mistakes, you have to do things right. These warnings sound pretty serious but I have to say I enjoyed the assembly, wiring and calibrating of this beautiful printer. When I printed the first good quality item, I couldn't stop watching it.
My first advice is to read more than one instruction because I found mistakes and gaps in all of them. Understand that each printer is different, nobody can give you a calibrated firmware, you have to do it yourself and that's the longest part.
I bought the kit from http://builda3dprinter.eu/, they offer good quality components, quick shipment (from Europe), and a website full of information. I found some gaps in their instructions, so I followed this manual too. The most common version of Kossel is the Kossel mini, but the one I've built is Kossel XL. The printable volume of the mini is a cylinder of about 170mm diameter and 240mm high. The volume of the XL is a cylinder of about 250mm diameter and 310mm high. The assembly is exactly the same.


Something I strongly recommend:

1 - Get an heated bed, your printed object won't move and you don't have add anything to the surface of your bed.

2 - With an heated bed, do not put your controller under it or it would get hot. I put it on top, in a case with an extra fan for it (it gets hot in Australia). I build the case with clear acrylic sheet (it can be shaped hot). The extra fan is from my CPU (I just upgraded my PC with an Arctic Cooler).


3 - Get a metal nozzle holder with heatsink: I melted the plastic one twice, before I ordered a metal one. I had to modify the platform and fan holder to fit the metal holder because it's much bigger. I also ground one side of the metal holder's heatsink to allow the probe to go up and down without touching it, because it's 25mm diameter (but there are smaller ones). Believe me, all this work is worth it.


4 - The fan that cools the nozzle holder must run all the time: I suggest you wire it up with the power supply rather then the controller, because if the controller shuts down, you don't want the fan to stop, causing the melting of the nozzle holder (or the extrusion platform in case you have a metal nozzle holder).

5 - When you troubleshoot a problem, if you are not sure of what it is, start looking for mechanical issues. Sometimes I look for the cause in electronics or software, when it's just something very simple, like a loose grub screw on the motor shaft (I spent few hours before I figure that out).

6 - PC power supplies are not always ideal. They are made to supply loads of different voltages: 12 V, 5 V and 3.3 V. If you only pull current from the 12 V rail, it goes down under 12 V, and it might not be enough to heat the bed to high temperatures. In my case it was 10.5 V and the heated bed was maximum 85 degrees, when I needed 100-110 to print ABS. So I applied a load (resistances) on the 5 V rail (which was giving 5.6 V) and the 12 V rail went up to 11.7 V. So I bought a LED strip power supply, which only have 12 V output and can supply 30 A (20 A would be enough). Now the voltage is constant and the heated bed keeps the temperature over 100 degrees.

(picture from ebay)

This is a video of my first print:


And this is the finished item: a bust of myself!


I've been scanned with a Sense 3D scanner, but this is another story... on my next post: 3D scanner!

These are other items I printed: