[Nails]

alphatroniqu -

If I understand you correctly, you don't completely place the part onto the paste, but rather just above it, and blow it onto the paste with a small low-pressure puff of air?

What size parts are you using when you observe this shifting problem?

[alphatroniqu]

Hi Nails

i affect any part more you have past more it shift , it just like put feet in mud
my biggest concern was whit 0603 and some 0805

on my system i lower part for just begin touch past then release vacuum
later i may try to implement the "puff" way but for now i not yet use it
but that part was still in developmet

Best regard

[Eclipze]

My understanding was that you need to push the part into the paste, but not squash the paste completely out the sides. Once the part has paste on it's pads, it avoids the part lifting or moving when the nozzle is retracted. I know from manually placing parts, if I don't get paste on both sides of a resistor, I risk having it tombstone.

[nisma]

My current use and understanding is to place the part at top of the solder, example pcb height + 4mil (printed solder mask is 6mil in this example) and then
add pressure to dissolve vacuum and have slightly positive pressure and going up with the nozzle.

[atlab]

... the text files aren't the way I need to go. I'm going to use SQL Server rather than text files to hold various tables of information. This lends a lot of flexibility...

Yes, the text format is good for commans via COM port and for projects (the text format allows to create and correct the project manually), but for other purposes i use SQLite - free and small.

[alphatroniqu]

HI

if i may help this is text file format i use for store my "Project information"
then my GUI translate back to modus ...

as for feeder it bank of 10 feeder 2 bank per side for left bank 1 and 2 = L1 ,L2
then L1-2 = bank Left 1 feeder 2

i also add modbus software / hardware comm

that not last file version but i may used as some base or use full for some one ?

Best regard

[Nails]

From what I understand from reading articles regarding placement force, Eclipze's description and reasoning is correct.

For larger components with thicker paste deposits, controlled height placement can be used and is sufficient.

For smaller components, a spring loaded head with an specific overtravel gives a known force according to the spring constant, which accomplishes the same thing - into the paste, but not too much.

I also remember one forum post I read a few months ago about a guy monitoring the vacuum sensor without a part in the nozzle and slowly lowering the head down to determine and set the board height.

[phomann]

alphatroniqu -

If I understand you correctly, you don't completely place the part onto the paste, but rather just above it, and blow it onto the paste with a small low-pressure puff of air?

What size parts are you using when you observe this shifting problem?

No this is not correct, for my Yamaha machine anyway. My machine only has a pneumatic Z axis, and it "rams" the part into the paste. There is no input for height.

Also the board needs to be will supported from underneath to stop it deflecting when parts are placed.

Cheers,

Peter.

[alphatroniqu]

Hi

Peter i assume your yamaha machine have some kind of slide tube on bit for control z force
on my JUKI machine outer tube was drive by machine and it have "Free" inner tube that slide inside it so pressure was only weight of inner tube + bit so ~20 gram
but that to much for be very reliable

so best was to control height of Z like all modern machine ..
also if you not control force or height you will risk to damage delicate part like MLCC cap
that may fail later in field

Best regard

[phomann]

Hi

Peter i assume your yamaha machine have some kind of slide tube on bit for control z force
Best regard

Yes, The nozzle is spring loaded and slides back into the head mechanism. The machine has a maximum component height of 6mm.

This was something I overlooked when purchasing the machine. It is classed as a chip shooter.

Cheers,

Peter.

[RomanLini]

Ouch. Bummer on the 6mm! Do you hand-place connectors or can you use a separate PCB holder that holds the PCB lower (and the pickup too I guess)?

Like a few of the builds here i'm using an RC servo for Z travel and have about 28mm of travel. That should be good for connectors and also other Z control like setting Z traverse height as for a lot of the parts it does not need to lift that high for the traverse, but on some parts and paths it may need to lift the full height to pass over other parts.

The obvious solution at this point was to code for a "set traverse height" command in the job file between pick and place commands. Using servo won't be as quick as your pnuematic Z but it's cheap and covers some of the other issues of a hobby level machine.

[phomann]

Hi,

Yes, I hand place the few components that are too large.

As to your 28mm clearance, that should be good for 13mm components, assuming that your lifting a 13mm component over one that's already placed. It gives a 2mm clearance.

I wouldn't worry about an adjustable traverse height. I can't see the merit of it. Just traverse at the full height and be done with it.

Cheers,



Peter.

[RomanLini]

Thanks Peter, I'm hoping to be able to place taller connectors than 13mm, like the larger USB connectors althoguh they are probably about 13mm anyway. Ideally a large part should never have a path that travels over another large part.

Part of the need for that much Z travel is because my PCBs are above the tapes, over the top of the tapes so it needs extra travel there. Putting the PCBs over the tapes means very short traverse distances for most of the discretes, to keep speed up. I'm hoping for about 3600 parts/hour if setup right, not because the machine is fast but because travels are small and efficient.

And the adjustable traverse height is mainly to compensate the slow speed of Z travel as the Z servos are a little slow taking maybe 150 to 250mS to travel 28mm, so you can see that is quite slow when added twice (for pick and for place) if full travel is used.

So with most discrete parts the Z travel and traverse height can be kept smaller to speed the machine up.

[Nails]

I hope the lack of activity on this thread means we are all hard at work with our designs and fabrication, and not that everyone has given up.

I am still wrapping up the details, but can now definitively say that I will be able to support 80 8mm feeders. 5 banks of feeders (each are 6 inches wide) on each side.

Here are some teaser pics of my almost completed feeder design.

[neilw20]

I received an email last night.


thanks for the info
PS: i got a present for ya! - A complete gantry off a pick and place machine, complete with suction head....

Can't wait to see what I have just acquired!!:wee:
Looks like I'll need to take some photos.

[RomanLini]

Photos! We need photos.

I haven't given up my PNP build but have needed to stall the build temporarily as I'm flat out with some commercial design work.

[Eclipze]

WOW Nails!!! That's looking quite serious!! Great to see!

I definitely haven't given up, but I've also got a load of work that has the priority over the hobby. At least for the mean time.

[Nails]

Thanks.

After a year of hardly any work, I too will have to take a pause for a while. I'm supposed to get a custom board tomorrow to do some FPGA work, which will last a couple weeks, then have a week long trip to NY for another company after that. Sometime after that, I am supposed to have a custom automated inspection machine to program from scratch which will take a few months.

Go figure. Almost finished, and now bunch of work. At least I am down to the more mundane things without having to think about designing anything tricky, so I should be able to pick it back up without much trouble.

Neil - PICS!

[neilw20]

With a bit of luck I will get it this weekend.
We might get some ideas from it. There will be plenty of pix.
If you are in Melbourne, (Vic, OZ!) you can eye ball it too.:cheers:

[alphatroniqu]

HI all

just for update the status of my project

machine was now working ;-)

ok like probably every project on this tread i end less
so it still have some tweaking to make and option to add
but machine was now able to assemble PCB ;-)

Machine doling HOME command then Self-Test then by 5 part PCB assembly and finish by home command

[nomedia="http://www.youtube.com/watch?v=FLT6qjUCdw8"]YouTube - Picture[/nomedia]

Best regard