What blow-off pressure for quick release

[Pilot65]

I'm in the initial stages of machine build (or rather design), have no physical parts at hand yet so can not examine or test anything.

What blow-off pressure should I expect to need for immediate component release... 0.1Bar... 2Bar... 10Bar?

I have no idea!

[Marek T.]

Many machine don't blow at all and release the parts. So any blow-off you will do it can be only still better than without it :-).
My machine blows and this is adjustable with some screw, but hard to measure the value as my sensors measure the vacuum values only...

[James Carter]

Hi Pilot,

I'm at the stage where I'm starting to put together some components, starting to build my pick and place head.

I'm planning on using a vacuum pump from SMC who have one with an analogue pressure sensor and blow off 'puff' which you can tune (with a screw). I don't think you need an awful lot of air, the paste as some decent sticking power especially on the light components.

www.smcusa.com/products/ZK2-Vacuum-Unit-Vacuum-Generator-Vacuum-Pump-System~90207

Happy to share my findings once I get the head out together.

James

[Marek T.]

Maybe I'm wrong but it's not vacuum pump but ventouri converter (ejector).
You need to supply it from the pressure pump, isn't it?

[James Carter]

Yes it is an ejector, I believe they call them generators when there are multiple stages and control electronics, sensors, blow off release etc.

You do need to supply air (approx. 10L/min)

[James Carter]

Ah, brain was not working, meant to say vacuum generator instead of vacuum pump in my first message.

[Mike Menci]

Marek,
You have solder paste on PCB - when the part is placed Athesion of paste is taking enough force to release the part from nozzle - as well larger parts are heavier and they will not stay on nozzles - they fall down easily.
Problems are mainly during testing where normally t there is no paste on PCB and many times parts do not stay on PCB.
Mike

[Marek T.]

Mike it's obvious for me. The thread was started by Pilot...

Yes James, this I've had on my mind :-). Just to be precise for Pilot that this module is not the only device that he need to create the vacuum, need pressure generator (pump) and this converter.

[SMdude]

When I was building my machine, I too thought that I would need blow off.
However, when I first got my machine going I only had vacuum setup and found that I didn't actually need puff at all.
I have my vacc solenoids mounted on in the head and it is a 2 way valve. The common port goes to the nozzle. The normally open is open to atmosphere and normally closed connected tut eh vacc line.
I then also though about how to make it even better as the 1/4" hose that I had from the valve to nozzle motor was too big and rigid. I ended up stepping the hose size down to 3/16". Now there is much less air that needs to be displaced each pick/place operation and I have never had a problem with parts sticking to the nozzle. Also, as the air rushes back in there will also be a small ram effect that should also help release parts.

If you don't run the solenoids on the head then you probably will need puff to quickly refill the line wih air each time you place.

Mick

[Brynn Rogers]

HI Pilot,

     I am also like James and will be using a vacuum ejector.   I have 24 of these, so a lot more than I can use.

They want 73psi, you power the 24v solenoid and it makes vacuum, closing an internal vacuum switch (the other two wires) then as the vacuum is released it gives a timed puff of air.

It has knobs to control the time and maybe pressure/size of the air puff.   I also have other little brass manifold like things that have adjustments to control airflow that can be added to the circuit if needed.

   I have the datasheets for the ones I have too, if you want to see it.

Brynn

[Brynn Rogers]

Also, the commercial machines use a puff of air when releasing the vacuum.    

[TheCunningFellow]

Commercial machines are aiming for as fast as possible I guess.

Although I have built the "puff" hardware for my machine but not yet integrated it into the software.

I found that as long as I have paste contact and a small dwell the parts will always stick to the paste rather than nozzle.

If I don't dwell for 0.2s then the vacuum can still very occasionally pull the parts up.

[SMdude]

Commercial machines also move much faster! That is a much bigger bottleneck with the current motion controllers,vs 0.1s extra dwell time at placement..

[Mike Menci]

Hi Brian,

There are very cheap vacum ejectors from Chiina- one here : https://www.ebay.de/itm/CV-10HS-Vacuum-Ejector-Generator-1-8-PT-Female-Thread-Air-Fitting-0-1-0-6Mpa/283117940101?hash=item41eb24ed85:g:jkgAAOSwylhbfPcV

Can sombody explain the operation ? What is needed to make this work?

Thanks

Mike

[James Carter]

Hi Mike,

The use the air from a compressor (about 80psi, 5-10L/min) through a nozzle to create a Venturi effect, creating a vacuum.

It means vacuum rise time can be very fast compared to a vacuum pump.

James

[Marek T.]

It's not 100% true. I have 6 Philipses with different firmwares and none of them makes any puff. When they are not sucking they are nonstop blowing.
It just depends on the ejectors types are used by constructor.

[Marek T.]

James you are wrong.
First, vacuum ejectors are not some amazing fast. They build the vacuum 0-100kPa near 80ms.
Second, the vacuum pump is normally nonstop turned on and the valves are switching vacuum for the nozzle, not the pump on/off. So vacuum building is as fast as valve fast is.
The pump fastness is not important at all.

My opinion is that Mick is right and this puff or blowing is not needed in so slow moving machines as are built usually here. Except the effect that it helps to keep the nozzle more clean inside.

[TheCunningFellow]

Yes - keeping nozzle clean sounds nice.  I just tried to use my machine last night an it would not pick.

It had dried past blocking the nozzle preventing any air flow.

Quick clean out and it's back to normal.  Puff could have helped slow that I guess.

[Mike Menci]

The idle, pick and puff looks like this per nozzle