3D DLP Printer, 14″ of travel.

hello,

is anyone use top down projection ? if yes did you have some probem during print ?deformation or anything else ?

Quote
A2

Print speed is determined by how many lumens your projector has, and the Z-axis mechanics, it can be as high as 50 mm/hr, but I think most are around 35 mm/hr.
I'm considering 26 mm/hr with a 0.10 mm Z-axis resolution.

One full step (no micro stepping) of the stepper motor advances the Z-axis:
.02623 inch/200 steps/revolution = .00013 inch, (0.00330 mm).

That's over kill on the Z-axis resolution, but I'm attempting to design a dead lift separation which will require a lot of force. This is being considered because I don't want to distort the object with tilting, sliding, or stair stepping the vat to separate the object from the vat bottom. I'm hoping I can flex the glass enough to initiate a peeling action. I'm considering using very thin Gorilla-glass.

The important factors in choosing a DLP projector are its native resolution, lumens, and contrast ratio.[/quote]


How did you get the ".02623 inch" figure? Also, what's a good z-axis mechanism if one was to build their own instead of going the brand name route? I would like to get a z resolution of 1micron if possible. Ideas?

Quote
NewKidOnTheBlock
How did you get the ".02623 inch" figure?

GT2 pulley circumference 1.57394 inch/ 60:1 ratio = 0.02623 inch, (.66624 mm).
0.02623 inch/200 steps/revolution = .00013 inch, (0.00330 mm, or 3.3 micrometer).

Quote
NewKidOnTheBlock
what's a good z-axis mechanism if one was to build their own instead of going the brand name route?

Linear ball bearings on round shafts just like a RepRap printer is probably your easiest, and least expensive route to take. I would incorporate a means to adjust the bearing preload. But with a 1 um resolution, you're going to want no play in your bearings with little friction, you should consider purchasing quality bearings. When you work with a high resolution machine typically they will use solid ways, or a positioning stage with roller bearings. You might need a temperature controlled environment.

Quote
NewKidOnTheBlock
I would like to get a z resolution of 1micron if possible. Ideas?

From what I have read usng a LASER ~0.001 inch is about the finest that the resin, and machine can mechanically and reliably print, but there are tricks, and mods that have been done to produce tiny parts, smaller than the pic below.

What do you plan to make?


Exercise: using a pulley, and a gear reduction of 180:1.

GT2 pulley circumference 1.57394 inch/ 180:1 ratio = 0.008744111 inch.
0.008744111 inch/200 steps/revolution = .00004372056 inch, (1.11 micrometer).

If you use microstepping you can reduce the gear ratio. With such a high gear ratio you should probably use a fine pitched metric lead screw, and gear reduction.

You should also calculate the stepper motor rpm, and torque to determine what kind of lifting force you are generating.

TORQUE, ROTARY TO LINEAR
[pic-designcatalog.com]

NEMA 17 Torque-RPM Chart
[www.pbclinear.com]

Torque Calculations:
[www.buildyourcnc.com]

Edited 1 time(s). Last edit at 05/31/2014 12:53AM by A2.

Hey, A2. I've been following your post regarding your DLP printer and was wondering if you could share with me how I could build a z-axis using v slot alu profiles, lead screw, and stepper motor. OpenBuilds.com has a 8mm acme lead screw (currently out of stock), do you think this would be strong enough and good enough to achieve 10micron layers?

Also, with regards to the VAT bottom what materials would be good candidates? I've read that Cortz glass and Borosilicate glass are both good. Would Plexiglass also be a good candidate? Are there materials that are UV transmissive that are flexible so a passive tilt system could be used?

Quote
NewKidOnTheBlock

Hey, A2. I've been following your post regarding your DLP printer and was wondering if you could share with me how I could build a z-axis using v slot alu profiles, lead screw, and stepper motor.

OpenBuilds.com has a 8mm acme lead screw (currently out of stock), do you think this would be strong enough and good enough to achieve 10micron layers?

Hi NewKidOnTheBlock.
I think you can get a better idea of how to design the Z-axis by reviewing the designs of CNC routers, check this link out for lots of examples: [www.mycncuk.com].

I've seen a small router using OpenRail with success, so I don't foresee a problem using the V-slot: [openbuildspartstore.com].

Search Google, and Youtube “Shapeoko gantry z axis”. Shapeoko uses V-slots, and you can copy what they are doing, and you will find many modifications.
[www.youtube.com]


To determine Z-axis forces you need to know the separation force (reference link below), the build area, motor torque, screw pitch, etc, and do a few calculations. I've provided a few examples for you to follow, change out the input values for your own.

Pulling force to separate the build plate from a vat coated with PDMS (silicone) can be found in this documentation. I used a force of 200lbs:

Digital Material Fabrication Using Mask-Image-Projection-based Stereolithography
[www-bcf.usc.edu]

LEADSCREW (POWER SCREW) TORQUE AND EFFICIENCY CALCULATIONS FOR GENERAL PURPOSE ACME THREADS (imperial, not metric)
[www.amesweb.info]


- Calculate the torque required of your stepper motor to lift 200 lbs with friction at the nut/shoulder pivot:

Note: If the motor doesn't have the necessary torque (plus start up friction) then you need to use a bigger motor or reduce the lead/pitch. As you decrease the lead (axial travel / revolution) you decrease the torque required to lift a load, i.e. your mechanical advantage will increase.

Torque required for an Acme-threaded power screw, (with friction):
[www.eng.auburn.edu]
Tr = (((F/cos a) x dm) / 2) x ((l + pi x u x dm x sec a) / (pi x dm – u x l x sec a)):

Variables:
F = 200 lbs.
a = 5.0 deg.
dm = 8.834 mm
l = 2 mm.
u = 0.08. (coefficient of dry friction)

Tr = (((200 lbs/cos 5.0 deg) x 8.834 mm) / 2) x ((2 mm + 3.1415 x 0.08 x 8.834 mm x sec 5.0 deg) / (3.1415 x 8.834 mm – 0.08 x 2 mm x sec 80.0 5.0):

Tr =
((( 200 lbs/cos 5.0 deg) x 8.834 mm) / 2) = 886.77444
x
((2 mm + 3.1415 x 0.08 x 8.834 mm x sec 5.0 deg) = 4.22871
/
(3.1415 x 8.834 mm – 0.08 x 2 mm x sec 5.0 deg)): = 27.75283-0.16061 = 27.59222

Tr = 886.77444 x 4.22871 / 27.59222

Tr = 135.9 lb-in


Torque Calculations:
[www.buildyourcnc.com]

Force to accelerate load, CNC Router:
[www.mycncuk.com]

NEMA 17 Torque-RPM Chart
[www.pbclinear.com]

Motor Sizing and Selection from First Principles - Torque Calculation
[www.festo.com]


- The following example demonstrates how much more force is required from a pulley drive compared to a very low friction ball screw drive.

Motor Sizing and Selection from First Principles - Torque Calculation
[www.festo.com]

(I filled in the values to the math that was posted on the internet):

Ball Screw, (90% efficient) / Lead Screw - The equation that relates the pitch or lead of the screw to the torque required to move it. Since the pitch or lead of the screw directly relates to the mechanical advantage the system provides, for low-lead screws i.e. 1mm, 2mm then the mechanical advantage is very high and vice-versa.

The equation is Torque (Nm) = Force (N) x ( Lead (m) / ( 2 x pi x 0.9 ) )

200 lbs = 890 N

Torque (Nm) = 890 N x (.008 m / ( 2 x pi x 0.4 ) )
Torque (Nm) = 890 N x (.008 m / ( 6.28 x 0.4 ) )
Torque (Nm) = 890 N x (.008 m / 2.512)
Torque (Nm) = 890 N x (.003185)
Torque (Nm) = 2.83 Nm (400.76 ounce-force inch) or (25.1 pound-force inch).


I have listed the units in the equation - the lead needs to be converted into meters since torque is usually measured in Netwon-Metres.

The '0.9' is the efficiency of the system. Generally speaking a ball-screw is approximately 90% efficient hence the 0.9 factor. In the case of a lead-screw then this figure can be much lower. I recommend 0.4 for lead-screw drives.

In conclusion, if you need more torque than you have try a shorter-lead screw, but don't forget to keep in mind the top speed of the screw and motor.

Belt and Pulley - The equation for belt-drives is much more straight-forward. Here the mechanical advantage is the diameter of the pulley used. This means in most applications a belt-drive is usually coupled to a gearbox primarily for the inertial match and then for the torque.

The equation is Torque (Nm) = Force (N) x Pulley Radius (m)

200 lbs = 890 N
Pulley diameter = 20 mm

Torque (Nm) = Force (N) x Pulley Radius (m)
Torque (Nm) = 890 (N) x .020 (m)
Torque (Nm) = 17.8 Nm, (2521 ounce-force inch) (157.5 pound-force inch).


- From what I have read, it's recommended that you double the required torque of the stepper motor.

Torque required for an Acme-threaded power screw, (with friction: 0.08 (coefficient of dry friction)):
Torque = (2174.4 ounce-force inch), (15.36 Nm), (135.9 pound-force inch)

Ball Screw, (90% efficient)
Torque = (400.76 ounce-force inch), (2.83 Nm ), (25.1 pound-force inch).

Belt and Pulley
Torque = (2521 ounce-force inch), (17.8 Nm), (157.5 pound-force inch).


- Note: as you increase microstepping the torque drops significantly, so you will need to adjust your torque accordingly.



Microstepping: Myths and Realities
[www.micromo.com]

Stepper Motor Dynamics Analyzer
[www.euclidres.com]

Microstepping Versus Gear Reduction
[www.ebay.com]

Microstepping to full step morphing is switching to full step mode when the motor is run at higher speed. It's a good plan as you get a bit less torque when microstepping.
[www.mycncuk.com]

CLOSED LOOP CONTROL OF STEPPER MOTORS WITHOUT POSITION SENSOR
[www.scribd.com]

Microstepping
[hydraraptor.blogspot.co.uk]


- To calculate the resolution of 10 micrometer you need to determine how you are going to drive the Z-axis, pulley, gears, or direct.

Example direct drive:
8mm pitch (lead)/200 steps/revolution = .04 mm, (40 micrometer).
8mm pitch (lead)/(200 steps*8 microsteps) steps/revolution = .005 mm, (5 micrometer).


You could also calculate the inertial/acceleration-Reynolds forces required.

Use this link to help with the conversions:
[www.translatorscafe.com]

You could also post your question in the buildyourownsla forum, I have a similar question posted here: Belt drive Z-axis
[www.buildyourownsla.com]

Quote
NewKidOnTheBlock
Also, with regards to the VAT bottom what materials would be good candidates? I've read that Cortz glass and Borosilicate glass are both good. Would Plexiglass also be a good candidate? Are there materials that are UV transmissive that are flexible so a passive tilt system could be used?

For the vat separation, from what I have read most are using a clear acrylic bottom coated with PDMS (transparent silicone). I've read that there is progress being made incorporating a Teflon film release, and it's probably going to be an improvement over PDMS, but it's still being experimented with. But I think you'll very soon be able to buy it.

If you can get Teflon film to work, you probably don't need a tilting vat. But from what I'm seeing most are using active tilting with PDMS, or jogging the build plate from side to side.

Thin borosilicate glass ~0.070 inch is what I have read researchers using because at that thickness there is virtually no UV filtering, but it breaks really easy. So it's not really an option for day to day use. But there is Gorilla-glass for $120 you can get a thin strong piece of glass. But again from what I have read, most are using acrylic because it doesn’t shatter like glass.

PDMS - Cliff vs Waterfall
[b9creator.com]


Best if you copy a machine that you like, has a following, and is some thing you understand.

I'm still researching, and learning about this technology, so do your own due diligence.

Edited 2 time(s). Last edit at 06/27/2014 01:00PM by A2.

Thank you A2!!

This post is a bit old, but in case someone else is also looking for the information
@arcidae2000 I designed and built a top down printer. Currently my wiper is switched off, because I need to redesign it. Printing small parts without a wiper isn't a problem, the problem comes in building big flat parts like a box. If you don't have a wiper the resin will dam up and you will have blobs in places. I you don't want to use a wiper you will need to angle the item, this means you will need supports and that costs more money and time to clean up afterwards. I have photos of the items I have printed so far with my design at gizmo3dprinters.com.au. I first wanted to show that my printer can print high detail. I am now working on printing bigger items. The idea is to print phasers for my laser skirmish business lasercommando.com.au. If I can print big and small items I think the printer will be ready for Kickstarter

Get sure to post in the crowd-funding section of those forums when you announce it please.

---

Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.

Quote
A2
Top Down DLP projector 3d printer

Ok, there is one negative with top down DLP printing, and that is the amount of resin to fill the vat.

Because you are moving the object down into the vat you are limited by the depth of the vat, and the amount of resin in the vat.
In addition you will need to make one small calculation to account for the volume of resin displaced by the armature supporting the build platform, that's not a deal killer just some thing to be accounted for.
So top down is reserved for small objects.

I wonder how degraded the resin becomes after printing an object, and how many objects you can print before the resin is set off.
If the resin degrades rather fast say after a few dozen parts, you could end up wasting a lot of resin.

But there is now UV resin costing only $22.50 + $12.00 shipping = $34.50/quart.
[www.solarez.com]

I imagine if you mask off the resin regions that you are not using you could further extend the resin pot life.

Could this problem be fixed with a cheap, higher density liquid underneath the resin that is withing the focal length?