Upgraded mechanics

[JuKu]

Besides, I saw the machine at the World Maker Faire New York last weekend. LitePlacer and ProtoVoltaics were placed on adjacent booths. We were so close that several people asked if we were the same company. I can assure the machine is real. I didn't see it actually moving parts, but my machine didn't do that either - both of us were running a dummy demo sequence to avoid reloading and cleaning up placed parts during the show. I have no doubt it works as advertised.

[mrandt]

I ask you please don't call us a scam without reason. KickStarters are a great way to gather market research.

Just for the record - I did not call anyone a scam. And as I take it WayOutWest's post, he was more expressing his frustration with Kickstarter (= Scamstarter?) in general than referring to your specific campaign.

So no hard feelings please

I think the "affordable Pick and Place" market is big enough for more than a few players.

I also believe there is benefit in exchange as there is much we can learn from each other.

The different approaches to CV for example... I understand you are using two (!) cameras in a right angle alignment (basically parallel to X and Y axis) tilted about 45° to PCB level. What I don't understand is how you eliminate the perspective error when measuring offsets, fiducial locations or component alignment. If you want to shed some light on this topic, maybe in a Blogpost on protovoltaics, I think we might understand better.

Btw, using a single camera orthogonal to the PCB (as LitePlacer does) also has many issues. Unless you have an expensive telecentric lense you will suffer distortion, perspective and projection errors, as discussed here to some extent:
http://liteplacer.com/phpBB/viewtopic.p ... rt=10#p351

[protovoltaics]

mrandt, sorry I was only referring to the comment about our video and nothing else.

Unfortunately, I cannot go into detail about our vision system.

[mrandt]

Fair enough. You will have to accept our speculations then

I guess LitePlacer will continue to focus on "straight down" camera anyways, as we can work around the issues I mentioned.

[JuKu]

Sideways camera for human inspection is not a bad idea. On the other hand, that doesn't need to be integrated into the software, it can work independently. Mrandt is right in that I don't have any desires to go into the math required for angled cameras.

[mawa]

I would like to see the 45° solution to get the machine fiducial coordinates when using fiducials I and many other pcb designers use: a 1mm gold/copper/silver circle within a 2mm green PCB colored circle with dark green solder mask producing the 2mm diameter with the same accuracy as a straight down cam does.

What about the new white solder mask some arduinos and other pcbs have?

Reliable fiducial measurement and the resulting compensation are IMHO a must.

Any proof we can see?

The only advantage I see in 45° is that glare effects are reduced as long the view direction is looking into the dark.

But calibration and measuring the cam-offset must be hell!! You never get exact 45° !

Looking at real pro machines I have not yet seen one with only 45° cams. Some have one to let the operator see the pcb placing pos up close because the pcb is "buried" in then center of the machine behind feeders and trays.

Well seeing is believing!

[JuKu]

If somebody is going to make a serious attempt at a magnetic nozzle addon for the liteplacer, I would definitely buy one.

Fyi, today I ordered a 3D printer and some Fuji nozzles.

[mrandt]

Fyi, today I ordered a 3D printer and some Fuji nozzles.

Yeah! Prepare to spend hours after hours with your new machine or just amazedly staring at it

Which printer did you end up buying?

To give you just two recommendations for 3D printing in general:

a) Get descent filaments. Both shape and raw material matter. Quality differences are huge and even the materials recommended by the printer manufacturer might not be the best.

Most recent innovation has not happened in the printers but in the plastic they print.

Melting and glass transition temperature, viscosity of melted material and especially warping / shrinking when cooling make materials easier or harder to print. Also the attributes of finished print differ a lot.

Find a nice overview here:
https://www.matterhackers.com/3d-printe ... nt-compare

And a good comprehensive writeup with explanations here:
http://nicklievendag.com/filament-guide/

I have used innoFill ABS a lot but recently switched to FormFutura ClearScent ABS for mechanical / machine parts - tremendous improvement. In PLA I found less differences. The strongest material I have used is Taulman Bridge Nylon - parts are somewhat flexible but virtually indestructible.

b) Whenever a print goes wrong, try to figure out why that happened and which parameters you need to tune. Over time you will learn a lot about thermoplastics, physics and your machine and slicer

A good starting point is this site - from the country next to yours:
http://support.3dverkstan.se/article/23 ... ting-guide

Good luck and have fun!
Malte

[JuKu]

>Which printer did you end up buying?
http://www.think3dprint3d.com/Kossel-Mini-3dPrinter-Kit

> Get descent filaments.

Their standard filament should be good: " the filament retains the typical features of PLA, but is tougher and less brittle. Due to a low shrinkage factor PLA will not deform after cooling." https://www.think3dprint3d.com/Filament ... A-filament

I also bought some magnets today. No points for figuring out what I'm thinking...

[WayOutWest]

[
Once I get my machine reassembled (see this) I will post a video that will blow your mind.

Okay, maybe not mind-blowing, but I finally got around to making the video: http://www.liteplacer.com/phpBB/viewtop ... 158&p=1012