New Attachments - Like a camera?

[a2retro]

If one wanted to design an attachment to replace the dispenser/probe for other purposes what would you recommend for prototyping?

Buy an extra dispenser and retro fit it or you have something else more suitable?

Glenn

[emcniece]

I'd love for this to happen, but Voltera would need to release their source code and document the probe API.

[jazozaya]

Hi @a2retro,

During production, we actually use a camera to calibrate and test that the V-One is indeed going to where it should be when aligning a PCB. This requires us to mount a downward facing camera on the carriage and have it move in XYZ.

For the longest time we had a frankenstein solution involving zipties and hot glue (ugly, but it worked), but a few weeks ago we cleaned it up and retrofitted the dispenser. The process was quite painless, we just needed to 3D print 2 small parts. The camera is now embedded inside the dispenser sheath and works great.

What type of attachment did you have in mind? Do you just want it to mount to the carriage and move in XYZ? Or you would you want it to actually interface with the gear or the probe pins?

[a2retro]

Ha, that was my first idea - camera for visual trace and solder joint inspections. Having it imbedded in the dispenser is a nice idea. I was also thinking of one that might be able to tilt so it would be attached to a small servo to get in on angles.

If you already have some way of integrating a camera back into your controller board so the video can be sent back to PC into your app, that would be pretty sweet but I'm assuming your just using a USB cam back to a pc app?

Glenn

[jazozaya]

Hey @a2retro,

You are correct - at the moment we just got a USB endoscope off Amazon and plugged it straight into the computer.

If there was a camera attachment, programming it to step through all the pads and take pictures for inspection shouldnt be very difficult.

I'm at home now, but maybe I can post some pictures of it tomorrow.

[a2retro]

I was considering picking up something like this - https://www.amazon.ca/Diosn-Waterproof-Borescope-Inspection-Megapixels/dp/B01JOE6W54/ref=sr_1_27?ie=UTF8&qid=1474605670&sr=8-27&keywords=endoscope+%2Bwifi

[deladriere]

Hey @jazozaya, could you share the 3D files for holding the camera and the reference to the camera. I have some vision problems and find really hard to adjust the printing head just in the middle of the pad's position.
I think a camera would help me
Thanks

[mewertow]

Hey @deladriere,

Unfortunately, I don't think that a vertical endoscope camera mount would help with aligning the probe. Removing the probe and mounting a camera would invalidate the tool calibration. This means that while you could mount a camera and find the pad center, this won't guarantee that the probe and dispenser will be centered as well.

You may have noticed that each time a tool is mounted, the printer does a short calibration routine in the X,Y, and Z axes. This is because of the variance that's introduced between different tools, or even between mounting the same tool twice - the position could have changed as much as 200 micrometers, which would definitely lead to alignment problems. This is also why we advise users not to move the tools after the XYZ calibration, or at least to release the motors and allow the system to re- calibrate.

We have another camera jig we use internally that snaps on to the dispenser and probe (which lets us monitor dispensing or probing up-close), but this requires a relatively expensive microscope camera which makes the whole assembly cost about $1,500.

• Matt

[deladriere]

@mewertow Yeah , this make sense
Thanks for your reply
Cheers,
Jean-Luc

[Eriond]

I think the camera is a necessity, first for visual inspection, but also as a component of a more advanced add-on: The pick-n-place module. Have a look at this link for some ideas what can be done on a shoestring budget: Liteplacer
Apart from the the cameras video back-haul (USB, I guess) and a stand-alone (USB-controlled) vacuum module, everything else is already in place in the Voltera. You even have the ability to rotate the vacuum needle using the syringe plunger gear that is normally used for paste dispensing.
What needs to be done, is to design a new attachment module that holds a vacuum needle with retractable tip, retraction sensor (connected to the alignment probe connector) and make that assembly free to rotate around the Z-axis by means of the plunger gear. Vacuum on/off is controlled separately via a USB device (the vacuum pump with electronic valve). Camera needs to be firmly integrated into the attachment assembly, so that its optical center offset is fixed over time, ie. doesn't require calibration.

I'd be happy to make such an attachment prototype. It would of course be much simpler if we could access a stripped down STEP model of the Volteras two attachment steel rods and other neighboring parts that create a physical boundary for a new attachment module.

[a2retro]

Eriond, While I am surely interested in a mini pick and place head for the V-One - where exactly would you envision picking/feeding the parts from?

[Eriond]

I understand your concern. The Voltera has a small area to the right of the probe alignment switches where I believe you could place 3-4 component strips next to each other. The carriage clearance is about 6,5mm, so if your components are higher than that, they won't fit. SMD parts tend to be small enough though.

Alternatively, or as a complement, you could utilize part of the reflow bed as component source area. That would of course impact on the maximum board size. All in all, I think it would be possible to have somewhere between 3 and 10 different components available in one pass. Once those have been placed, the pick-n-place application would home the head and wait until you manually replenish the component stack. Then the process is iterated.

[mharris]

You can build a machine for OpenPNP for under $1k, a pretty nice one for $1.5k. The V1 is way too small for PNP use... with a $1500 OpenPNP machine you'll have dozens and dozens of strip feeders, two pickup heads with auto tool changing, Many people have totally reliable machines running openpnp now, I know of several who are populating dozens of boards every day with theirs.

[Eriond]

After having a closer look at the OpenPNP project and the various machines built from that base, I see that a Voltera V1 PNP add-on would probably be not so useful. I still believe that that you could build a great PNP attachment with all necessary features, but if there isn't enough area to source components from, the whole set up becomes pointless.

Maybe, and that's a big maybe, if you would build a complete PNP + Component feeder module, then it could be useful. But that's a completely other beast were talking about, not just a "simple" carriage attachment. (And that thought kind of took my enthusiasm away. Hopefully it sparked someone else's)

[mharris]

The OpenPNP community is wanting to use the V-One's dispenser and mount it onto the pickup heads to be able to deposit paste onto boards, like an Essemtec Scorpion would. I don't see that happening any time soon, there isn't a lot of drive for it (since stencils are so cheap and easy). I was thinking of working on that, but instead I bought a full V-One haha.

[KikoLobo]

Do you happen to have the 3D printed part STLs?

[Icky]

+1 on the STL’s for the printed parts but also any pictures of the setup would be helpful

[Pescatore]

@jazozaya
Would you mind describing how this works? It seems that you were only watching the video feed to see where
the probe goes. However, once you remove the modified dispenser, the alignment is no longer valid for the real dispenser.

[a2retro]

Getting a model of the dispenser housing or even a dimensioned diagram, would be helpful as I would still like to make an external endoscopy camera holder that clips on the dispenser with additional magnets.

[froginham]

If it is of any interest, I created a 3D printed camera mount which slides on to the top of the carriage which uses a usb fibre optic camera which greatly helps when aligning boards and inspecting during the print and solder pasting processes. Attached are some photos.IMG_1954.HEIC (1.8 MB)
IMG_1955.HEIC (1.9 MB)