SMALLSMT Vision Placer Pick & Place machines

[mrpackethead]

Im curious, does the machine in the video have the ability to place a  0.4 pitch mm TQFP?   You'd need the machien to be able to place that with about the same level of accuracy as an 0201? Does this mean if you want to place those parts you need to run it that slow as well?

[Smallsmt]

First of all again this VP-2500DP (6400usd) machine is not recommend by us for this job.

Im curious, does the machine in the video have the ability to place a  0.4 pitch mm TQFP?

I think it could be possible but it's not reliable you need to adjust a lot.

If you want to place 0,4mm pitch you should buy HP machine from us.
We offer VP-2500HP, VP-2800HP, VP-2800HP-CL22, VP-2500HP-CL32, VP-2800HP-CL64 to do this job.

Does this mean if you want to place those parts you need to run it that slow as well?

The precision placement of fine pitch parts need to slow down Z axis speed on mostly all pnp machines on the market without force feedback.
For the belt machine the X/Y axis speed need to slow down too or use a long acceleration ramp.

Finally we are selling prototyping and small production machines.

If placement speed has the highest priority you need to buy a big machine (100K new or used) and spend more money to receive high speed.

Please compare the working speed of 0,5mm TQFP placement in this video using a VP-2800HP-CL64 (14950usd)

https://youtu.be/9M92HoxJYpo

[Smallsmt]

Here you see VP-2800HP-CL22 machine placing a TQFP208.
The Z speed is reduced during placement.

[Kjelt]

Please compare the working speed of 0,5mm TQFP placement in this video using a VP-2800HP-CL64 (14950usd)

The linear guideways construction keeps amazing me. Single unsupported rail on the left, dual rail with ball bearing in the right. Never seen this type of asymmetric construction before but it sure seems to work.

[rx8pilot]

The gantry looks rather lightweight so breaking the linear way design norms is not too tragic.

It does look crazy.

Short and misplld from my mobile......

[Smallsmt]

Please compare the working speed of 0,5mm TQFP placement in this video using a VP-2800HP-CL64 (14950usd)

The linear guideways construction keeps amazing me. Single unsupported rail on the left, dual rail with ball bearing in the right. Never seen this type of asymmetric construction before but it sure seems to work.

Sometimes it funny to see people writing comments about mechanical constructions done by professional mechanical engineer.

Yes this works perfect!
One unsupported rail is able to carry 100Kg or 250Kg supported.
Our gantry  is very rigid and stable.
It has dual guide lines.
The head construction weight is around 3Kgs.
We use 1000W AC servo for Y axis on VP-2800HP-CL64 and 200W on X axis.
The VP-2800HP-CL22 has 400W AC servo on Y and 200W AC servo on X axis.

[mrpackethead]

First of all again this VP-2500DP (6400usd) machine is not recommend by us for this job.

You'd need the resolution for a .5mm QFP as well.    I'm wondering, why did you even put the video up?

[Smallsmt]

It's a customer video Alexey Volkov uploaded the video.
Ii thought it's cool to see what's possible even on a simple machine.
He did his own modifications for part trays and PCB holder.

[mrpackethead]

It's a customer video Alexey Volkov uploaded the video.
Ii thought it's cool to see what's possible even on a simple machine.
He did his own modifications for part trays and PCB holder.

Yeah, its quite amazing what can be done for a few thousand dollars in china.

[sigmut]

Thanks for interest to my 0201 project assembly using SmallSMT's simple machine.

This is 20x10mm pcb with 22 components (mainly 0201 + 0.5mm pitch 81 ball BGA) on top and
45 components (mainly 0201, X2SONs and two 0.4mm pitch connectors) on bottom side.

Their machine is very rigid, uses good motion parts and allows to get enough accuracy to install small quantities of 0201.

Of course the speed is very slow. As I remember, I spent aproximately 30-40 min. to assembly both sides.
But I was need only 10 pcb prototypes for my customer and speed is not issue for me.

In this project I ran into solder printing problems instead of machine problems.
I used 0.08 stainless steel stencil, but still very hard to achieve good printing through triangle apertures for X2SON.

So real assembly of 0201 with SmallSMT's stepper+belt machine is possible. This pcbs are working fine without any problems.

[Kjelt]

One unsupported rail is able to carry 100Kg or 250Kg supported.

The weight that can be carried is not what is important here.
The amount the rail bends is what is important.

[rx8pilot]

One unsupported rail is able to carry 100Kg or 250Kg supported.

The weight that can be carried is not what is important here.
The amount the rail bends is what is important.

FWIW - I am an engineer that has designed machines with linear rails and I was told never to load a rail unsupported. I have never seen any machine go against this rule of thumb - even for light loads. The unsupported load 'capacity' is subjective as well as having important variables. The length of the unsupported section and the distribution of the load as well as the system tolerance for flatness.

That is not to say that this configuration is unworkable at all....it is unlikely to be a significant source of error relative to the other contributors. It is, in my opinion, a bit of a red flag that the engineer is willing to break the rules in this way. With that said - it appears to be a better overall solution than other desktop systems.

[IconicPCB]

 have seen other machines of the gantry style of construction using unsupported linear bearings on one side of the gantry while the gantry is either cantilevered unsupported beam or simply supported beam.

In fact i used to own a flying tester probe where the bearing a round rod was unsupported and the other end of the beam was a simple slider without a bearing block to locate it.
There is a German pick and place machine which relies on a a simple cantilever beam design.

The square rail linear bearing  is specified in a particular way in order to accommodate the forces usually associated with typical burdens found in metal working applications. 

[SMTech]

There is a German pick and place machine which relies on a a simple cantilever beam design.

Which would be a Fritsch, last time I was looking the price on one of those was the same or a bit more than a faster machine from Essemtec or Dima, IIRC it was the non-std options that killed it for me. Still it seems to work for them, the current models seem to be broadly similar to 6-7 years ago.

[IconicPCB]

No,

It is a Mekatronika model not to be confused with a Mechatronika from Poland.Unfortunately can not locate info presently.

[Koen]

FWIW - I am an engineer that has designed machines with linear rails and I was told never to load a rail unsupported.

It's on page three of any linear guide datasheet : "the rail isn't straight/correct by itself, you have to tie it to a reference".

That being said, it's nice, it's light, it's smart and it works. Well done.

[Kjelt]

But an extra aluminium bar under it with a few extra vertical supports would have made it much better.
The penalty would be two less feeders.

[Koen]

Why ? What's the issue here ?

[Kjelt]

The rails are used outside the manufacturers recommendations, what the consequences will be can better be asked at the manufacturer.
My guess is that the forces and stresses on the wagons and balls will probably be higher when the rails are unsupported and can bend.

[Smallsmt]

You’re wrong the guide lines take less load no need to support the rails to reach our precision. And it’s not out of manufacturers specifications! Totally useless to say they can‘t bend because the force applied to the rails is too low!

[SMTech]

No,

It is a Mekatronika model not to be confused with a Mechatronika from Poland.Unfortunately can not locate info presently.

http://www.mechatronic-systems.com/products/smt%20placer/P20_e.html & Fritsch both do it The Fritsch head looks a little heftier, I wonder if either system has to adjust the acceleration or path to counter the effect of the head at the end of the cantilever, the Mechatronic in particular is very slow. Mechatronic lists two supposed UK dealers neither of which list them or ever have to my knowledge, (well maybe the printers).

This is certainly not a common approach, most machines in fact would support & drive the gantry from both ends, whatever the rating you just can't deny physics and forces do not do right angles.

[Kjelt]

You’re wrong the guide lines take less load

Less load than what?
It takes the same load as the opposite side where for some weird reason they decided to use double rails.

they can‘t bend because the force applied to the rails is too low!

Looking at the length it is safe to say it will bend but probably a negligable amount.

[Kjelt]

You’re wrong the guide lines take less load

Less load than what?
It takes the same load as the opposite side where for some weird reason they decided to use double rails.

they can‘t bend because the force applied to the rails is too low!

Looking at the length it is safe to say it will bend but probably a negligable amount.
Place a video with a micrometer dial set in the middle to the lower side of the rail when the bridge is on the far side and move the bridge to the middle, look at that needle go......

[mrpackethead]

You’re wrong the guide lines take less load no need to support the rails to reach our precision. And it’s not out of manufacturers specifications! Totally useless to say they can‘t bend because the force applied to the rails is too low!

Your wrong.   Wrong in respect of how to market your products.     referencing this video, was just asking for you to have the machine criticised by evryone because it doe'stn do a particaular job very well.  Your messaging is confused. 

[Koen]

It's a 6kUSD benchtop machine placing 0201. What the hell are you on about ?