[Robin Hewitt]

Robin- Thanks for the tip on the sensor. Farnell price is $96 for a single sensor(!) and 0-15 PSI is not a good range for me. I would prefer 0-3 PSI or similar range to try and detect the component.

Blimey, I got mine for GBP 12.23 from RS last month but they've jacked their price up by a fiver since then

I went for one atmosphere pressure, but 3psi should be plenty enough for pnp. They Honeywell range has 1psi and 5psi, didn't see a 3. The hysteresis is .15% of the sensor span so my 15psi part is okay for any vacuum.

I was planning on a vacuum reservoir with the sensor deciding when to turn the pump on. Didn't seem like a good choice for pick up detect but could be worth a try. I suppose I should look for the vacuum decay rate with the pump off

[Nails]

In case this is valuable to anyone, this is what I am going to use for my vacuum generation. Vac Pro

It uses shop air and provides a high vacuum with high flow rates. I think most commericial pick and place units use something like this.

If you listen to commercial pick and place units, you can hear a fairly loud hiss of air at the time of picking the part, which I think is probably the compressed air exiting one of these devices to create the vacuum. I doubt they use aquarium pumps or similar.

[johnohara]

Interesting thread. As is RomanLini's.

The PnP machine shown in the SparkFun video is from Manncorp.

Here's a link to the MC-384. Under the "specs" tab, at the bottom of the page, is a reference to the vac specs. 4 CFM @ 80 psi equates to the Vac Pro Plus.

I'm not affiliated with MC.

~John

[Nails]

Nice find.

Unfortunately, I took a stab in the dark earlier this year and have a vac pro basic, which produces 2.1 CFM @80 psi.

Wouldn't be the first, or last time I bought something then changed my design. Or the hundredth time, for that matter.....It is so tempting to buy parts before the design is complete to have something physical. I'm learning this lesson the hard way.

Just when I think my design is close to being finished, "feature creep" changes everything. I am sure I am preaching to the choir, since everyone is farther along than I am.

[johnohara]

It wasn't my intention to discourage you or make you feel as though you made a mistake. I apologize if I did.

My guess is that the MC-384 is probably "over-spec'd" in its vac design for a number of reasons.

Its servos can position the vacuum tube over the top of a chip very quickly. My guess is that the engineer(s) responsible for determining how much vac is needed just to keep up at that point probably erred on the side of caution, borrowed from a previous spec, or were forced to use a vac employed across their entire product line.

Voila, 4 CFM, or more.

Those little chips don't care how much vac is being pulled so long as their ride is firm, fun, and accurate.

CFM does contribute to placement accuracy. If the chip doesn't shift in transit, tighter placement tolerances can be achieved. If you're the person tasked with deriving the optimal CFM to grab a chip quickly, hold it as tightly as possible, then release it as accurately as you can with zero post-positioning, you'd over-spec it too.

2.1 CFM should work fine. Why wouldn't it?

~John

[Nails]

Oh, not discouraged in the least. I'm sure the lower vacuum rate would work as well. May just need a few 10's of milliseconds delay before moving to allow full vaccum, is all.

It's just easier to emulate someone's already known working specs for peace of mind. You have one less variable to consider when you are wondering why a part might be falling off the pick-up head.

[Eclipze]

I want more time to play!!!

[johnohara]

Eclipze:

If it's not too much of an imposition, could you post a few more pictures of your base?

I note in one of your recent posts that you used Corian. I'm assuming that was because it was easier to shape and work with.

Last spring, I had great success creating Epoxy Granite using aluminum oxide. With exception to air entrainment, which can be fixed with vibration, I believe what I accomplished would work exceptionally well with this project.

Here's the post: http://www.cnczone.com/forums/755168-post16.html

~John

[Eclipze]

There are a couple of pictures on page 17 of the base. There isn't much too it, other than the holes, slots and hex pockets on the underneath that hold the nuts captive for the PCB area (bolt area). The edges are also stepped on the underneath side, as it's made to fit in a square tubing frame.

I chose corian because it's stiff, flat, nice and heavy, not effected by humidity and it's rather easy to machine

[RomanLini]

I wouldn't get hung up on that "4 CFM" spec. The pick and place doesn't need CFM, it needs a vacuum. They just use the shop air compressor's 4 CFM in a crude venturi device to generate some vacuum and unfortunately they don't state how much vacuum or the figures might actually be of use.

I am using an aquarium air pump. It has worked well for me for years in hand placement and makes enough vacuum for PNP provided the forces are not abused and the right nozzle is used for the part. The aquarium pump is quiet, energy efficient and fully tolerant of continuous use so you don't need to mess with large vacuum resevoirs and on/off switches etc. Just the pump, and a 12v SPDT solenoid valve.

If you use a higher strength vacuum you can get more hold on the part, so you can move the machine quicker but then it's harder to drop the part and that takes time (slowing the machine down) or it needs complexity of regulated air puffs etc to blow the part off the nozzle, and maybe cameras to see the part position after a high vacuum has moved it (when the vacuum was applied). I did all that testing years ago and I like a low vacuum system. Also, once using rubber cups for larger parts a high vac becomes a bad idea as it deforms the rubber giving more part deflection.

I hope to get 2000-4000 parts placed per hour using just an aquarium pump and good machine design including short traverse distances, good nozzle choices and fast nozzle changes. Increasing complexity, cost and noise just to bump that speed up a bit higher doesn't interest me, the bottleneck will probably be my stenciling and cooking anyway.

Eclipze- What's Corian? Is it a polyacetal? Do you know the actual material name? I have some black HDPE here for some of the next parts I need to cut but it's a little softer than ideal. Thanks.

[phomann]

Hi,
Corian is made by Dupont. It is used for kitchen bench-tops. It is very expensive unless you have a free source of off-cuts.

DUPONT CORIAN SOLID SURFACE

Cheers,

Peter.

[Eclipze]

Corian is a common material used in making kitchen benchtops. It's expensive to buy, however often kitchen cabinet makers have a lot of off cuts they need to unload ;-)

It's a material made by DuPont.
Welcome to the world of DuPont? Corian®

"DuPontTM
Corian®
is a solid, non-porous, homogeneous surfacing material, composed of ±1/3 acrylic resin (also known as PolyMethyl MethAcrylate or PMMA), and ±2/3 natural minerals. These minerals are composed of Aluminium TriHydrate (ATH) derived from bauxite, an ore
from which aluminium is extracted."

It's very hard, resists compression, but can be brittle with thin sections. It's a really nice material to machine, although it generates a lot of mess hahaa...

I concur with the vacuum vs CFM. I want to have a solenoid on the carriage to control the airflow, so the volume that is put under vacuum and released from vacuum pressure is very small.

[phomann]

I want to have a solenoid on the carriage to control the airflow, so the volume that is put under vacuum and released from vacuum pressure is very small.

Yes you need the valve as close to the pickup head as possible to minimise any lag in the vacuum system.

Cheers,

Peter.

[Robin Hewitt]

If you use a higher strength vacuum you can get more hold on the part, so you can move the machine quicker but then it's harder to drop the part and that takes time (slowing the machine down) or it needs complexity of regulated air puffs etc to blow the part off the nozzle,

You get an automatic pressure pulse when you crack the vacuum. If it's too limp increase the bore of the vacuum hose. Make sure the valve is fast switching and full venting to the atmosphere. If you have a cavity above the nozzle consider a conical hole feeding the tip.

[Robin Hewitt]

Yes you need the valve as close to the pickup head as possible to minimise any lag in the vacuum system.

Gosh I'm getting all controversial

A longer pipe will have more moving air in it that has to be stopped at the nozzle. It just needs to be big enough diametrically so that the air can remain vaguely stationary where it contacts the tube walls and rush in down the middle.

[RomanLini]

Thanks for the info on the Corian! I'll look for some. Does it blunt cutting tools being 2/3 composed of aluminium trihydrate?

I'm interested in anyone's vacuum designs (Eclipze, Peter and Robin).

My tests used a 12v SPDT air solenoid. SPDT means it's a changeover with one common pipe that switches flow between two other pipes. I put the middle pipe to the pickup head and the other pipes to the vac pump and also to atmospheric. It worked well with hose length from 6" to 2 feet from the pickup head to the solenoid. Pretty much instantly getting the vacuum at the head, and when the solenoid switched off it made a fast pressure pulse back that released the part well.

I didn't think tube size was that big a deal, as the vacuum pump only makes a couple of PSI total, it only needs to evacuate about 10% of the air in the tube to make that -2 PSI and fully establish the vacuum.

To my mind it would be best to keep the tube short and thin, I planned to mount the solenoid on the moving gantry and keep the tube about 250mm long but I am keen to hear your thoughts on this Robin and any suggestions in general.

Peter- Do you own a commercial machine? Is there any experience you can share there?

[Eclipze]

My main concern with the solenoid on the gantry is mechanical switching movement creating movement at the tip at pickup. To help avoid unwanted movement, my intention has been to align the solenoids plunger vertically, as well as provide a soft mount to absorb the shock.

[neilw20]

You need something small !?? This may help.
Ask about a DR120 valve. I could not find online. It weighs 28 grams (1oz) with metal air fittings.
Check this out
http://tpcpage.com/products/pro01.asp
I asked nicely, and free of charge they fedexed me a catalog .. maybe a 1000 pages at a guess

[Eclipze]

Very neat!

It always comes down to availability and cost. Could find much with a search. A local supplier would surely charge $100 for it. I have an idea on how to make my own part

[neilw20]

There is no local supplier.