[James Newton]

IMHO = In My Humble Opinion.

Interesting idea using the nozzle vacuum to sense the position of the board.

[rgee]

re fliperspace question on component rotations
there is an IPC standard see links

http://landpatterns.ipc.org/IPC-7351...Convention.pdf

http://pcdandf.com/cms/images/storie...02hausherr.pdf

PR

[flippersplace]

Thanks RGEE exactly what I was looking for. Now to go and edit my footprint library..

[Eclipze]

rgee... great stuff!! Appreciated. Never looked for naming convention info, mostly because I dread the thought of renaming and loosing sync with all schematic components hahaa...

[flippersplace]

Selective Soldering Machine Jig Making Process

[phomann]

Selective Soldering Machine Jig Making Process

I didn't even get a chance to start my own thread.

actually, I'm about to embark on converting it to being controlled by Mach3. It should be an interesting exercise.

Cheers,

Peter.

[flippersplace]

I've had very good results populating Excel then having VB Script generate all the GCode for Mach 3

[rgee]

Hi there guys
There has been quite a lot of discussion in the top 3 pnp threads regarding drilling holes through stepper motor shafts.
I have the ability to do this but chose not to for a few reasons. So I thought I would share how I made my head, my aim was not sub light speed or even the speed of sound, My z axis is small, it was made out of an old hard drive head mechanism and has been used as a holder for a pencil die grinder for a cnc engraver. The other reason is I felt that you do not need the power of a larger stepper to rotate a small smd part. There is also the inherent problem with joining two shafts and keeping them concentric.
The basics head fits in the sleeve where the die grinder went, this has a diameter of around 18mm.
The head features the ability to rotate a full 360 deg. It Is also spring loaded with a travel of about 8mm.This is achieved with telescoping square brass tubing. My future aim will be to have multiple z axis units with different nozzle sizes, this would also allow for parallel picking. I believe that putting my energy into duplicate z axis will give me more flexibility than a nozzle changer. For the majority of the parts likely to be placed by the backyard technician 3 nozzles should suffice.
parts as labelled
A
this is a piece of delrin 6mm one end to fit the silicone tube to the vacuum ejector.
As my machine had a pneumatic die grinder and solenoidal value I just run the vacuum ejector
off the spindle on signal to create vacuum.
B
This is the sleeve that fits my z axis holder it has a hole all the way through and is fitted with 8X12mm bearings at both ends. These are glued in place with thread lock. Bearings are from old photocopier.
For those who do not know old photocopiers are a great source of parts for cnc'ers / hackologists
C
This is a sleeve that couples the 5/16 square shaft to the bearing in the top also to the drive gear
D
plastic 75 tooth gear from same flatbed scana as motor
E
This is a collar that stops the centre shaft from falling out, held in place with a 3mm grub screw
F
This is the vacuum delivery tube it is silicone fuel line for a model engine found in hobby shops.
Internal diameter of 1/16th ,the length is enough to allow over 360 deg rotation.
G
This is a short piece of silicon tube placed over the end of nozzle for picking large parts, any thing from tant B to QFP.
H
spring to give nozzle an override / travel of about 8mm.
I
delrin bush to keep spring centred on shaft.
J
Aluminium bush to fit square 5/16 square shaft to the front 8mm bearing this needs to be drilled to a diameter equal to the distance across the corners , made to a tight fit. If you do not have a set of fractional drills I recommend you add to your wish list as they are invaluable
K
Square 5/16 brass box section this is from hobby shops K&S metal products there is a large selection of profiles, I had square to hand so that is what I used, there is also hex profiles that would also suit.
Most of the profiles telescope making them ideal for this task.
http://ksmetals.com/resources/cuttolength.pdf
L
This is the business end of the nozzle the end has a .6mm hole with a reduced end to 1mm, this was used to descend into the part pocket to drag the tape forward before picking the next part. This was turned on a small lathe to fit the end of the 3/16 shaft with a flange to support the spring. This was soldered in place.
M
3/16 brass box section this slides inside the 5/16 box section. To make it slide easy I needed to polish the outside with 400 grit paper. When purchasing it would be wise to check that the two will slide together and have no bend. The top end of this shaft has a small tube soldered in, to allow the connection of the silicon tube.
N
stepper motor 72 steps per rev recovered from a flatbed scanner the pinion and plastic gear were part of scanner. To do one revolution with a gecko251, I need 35000 steps this is more than enough resolution. A full rotation takes a fraction of a second.


That's about it for now
Rgee

[RomanLini]

Very cool! Thanks for posting that.

It's nice to real a post about real hardware for a change, compared to the "general discussion" posts... OK, the general stuff is good to read too, but seeing photos of what someone has built is a treat.

[Eclipze]

I really like you idea of having 3 placement heads, and no nozzle changing system!
Quite practical.

[flippersplace]

Ok, finally managed to pull this footage off my slowly dying camera.. from my August conversion from the DVD head mechanism to pneumatics..

Video editing is a bit rough near the end but was finding it tedious.

Intention was to try get experience driving the pneumatic heads as an option for the eventual Mk 2 version. These are second hand one way Festo micro cylinders mounted on micro linear slides

I like the idea of not needing precise Z positioning with the use of air cushioning to replace the previous spring mechanism. Had/have some issues with stiction across the various 8 heads (all identical but all second hand with various wear).

Initially I had to use very low pressures to try to balance down pressure on components vs speed but the heads were not were not returning fast enough so added external return spring so I could up the supply pressure.. Might play with some 2 way cylinders in the future if I can get hold of some small enough..

Will do some more experiments with heavier springs / more pressure in the future.

[ame=http://www.youtube.com/watch?v=VQLev5tI64Q]DIY Homemade PCB SMT Pick Place machine 2011 August - YouTube[/ame]

[RomanLini]

I didn't even get a chance to start my own thread.
(re; selective soldering machine)
actually, I'm about to embark on converting it to being controlled by Mach3. It should be an interesting exercise.
...

Peter- Do you have any more photos or info on the selective soldering machine? I often pay to have small quantity manufacturing done (outsource it) and a machine like that would mean I can do some production work here on the machine (rahter than hand solder or outsource).

How does the solder feed? Can you post any photos of the head and solder feed mechanism?

Sorry Flippersplace for going off-topic in your thread!

[stewi]

Have you considered moving the board over a mini-fountain instead of solder iron and solder feed?
[ame=http://www.youtube.com/watch?v=qPMij0_2RO0]Selective Solder Machine - YouTube[/ame]

[flippersplace]

Sorry Flippersplace for going off-topic in your thread!

No its all related. I have vast quantities of TH jacks on all my boards that I hand solder vast quantities of so I have an interest in a soldering robot as well.

It's another area on my list to automate in some form so I'd been keeping an eye on it hence my jumping on Peter's vid when it came up.

[phomann]

Peter- Do you have any more photos or info on the selective soldering machine? I often pay to have small quantity manufacturing done (outsource it) and a machine like that would mean I can do some production work here on the machine (rahter than hand solder or outsource).

How does the solder feed? Can you post any photos of the head and solder feed mechanism?

Sorry Flippersplace for going off-topic in your thread!

Basically, the solder is feed off (pulled) the reel by 2 knurled wheels that feed it down a tube (teflon?) It then comes out near the soldering tip.

The setting include a;

• solder prefeed

• preheat

• 2nd feed once the tip is in contact with the joint

The feeds are programmed in as a distance in mm. But, that's just a calculation done for timing of the motor driving the knurl wheel. The 2nd knurl wheel is spring loaded to put pressure on the solder being fed between them.

In addition to the solder being fed to the soldering tip, there is also a air hose as well. It delivers a bust of compressed air onto the tip to clean it by blowing crud, slag and excess solder off the tip, when commanded. It is one of the standard soldering cycles. I use it before each board and fter about every 50 joints.

I will take some more photos, write a bit more and put it up on the blog this weekend.

Selective Soldering Machine Jig Making Process

Selective Soldering Robot Conversion to Mach3

Cheers,

Peter.

[phomann]

Have you considered moving the board over a mini-fountain instead of solder iron and solder feed?

A CNC solder fountain selective machine like this is my ultimate goal.
Wave Solder Machine, Selective Soldering Machine - Q & A

It even runs off Mach3. They had a cheaper/simpler model when they first started out.


I just need to find a reasonable solder fountain. There are ones like this;
CM2720 Titanium Alloy Jet-flow Solder Pot Bath 4KW 220V | eBay

But the form factor is wrong as the motor sits higher than the fountain. It's a little expensive to experiment with as well.

Also, you need to set up a fluxing system and possibly a nitrogen supply to keep the dross away.

Cheers,

Peter.

Cheers,

Peter.

[rgee]

Hello Peter
careful what you wish for, I spend a lot of time swearing at a selective soldering machine.
you already have a soldering solution I doubt you would gain any speed from a selective wave.
The only advantage I can see over the wave process is if you are working on multilayer boards.
A few disadvantages
long start up time, got to get the whole pot up to temp
cleaning out the pump.
Cleaning dross out of pot, yes even with nitrogen you still need to clean every 40 hours
fixing the bridges and other problems created by a wandering wave.
They are not so good on densely populated boards.
Having said all that any good tool box need more than one type of hammer.
Just my 2cents worth.
A substitute for nitrogen could be CO2 this can be made by fermentation, 2liters of sugar water and yeast will make 30 litres of CO2.
A two nozzle selective soldering machine uses 2.5 litres per minute.
Rgee

[RomanLini]

Thanks for the info Peter!

Stewi- I prefer the idea of a top down soldering station to the more moden micro fountain type under the PCB. I think top soldering uses gravity to help wet the solder down into the PTH hole, and it's easier to hold the PCB that way up in a fixed jig rather than hold move it around over the top of a tiny wave spout (where you can't see it).

It's better for inspection while running, and you could even do some touch ups on top of the PCB while it's still soldering on top of the PCB.

I look forward to seeing more of your top-down selective soldering machine Peter! Unfortunately google only seems to bring up wave type selective soldering machines these days...

I have some hollow point soldering tips, that were used in an old electric solder sucker iron (that never worked that well bt the way).

What about using a hollow tip for the top-down soldering, and feeding the solder in above so the wet solder comes out the end of the tip onto the joint? That might work well.

[rgee]

Hi all
there are some tips for soldering robots in the JBC range
Cartridges and tips - JBC Soldering and rework tools for electronics
down the bottom of page
Peter R

[flippersplace]

So, while I'm really really excited about Stewi & colleagues Wikiplacer project I had already been underway with my second pick place machine.

The biggest difference, after using (read nursing) my previous model is that reliable automated feeders where possible is a must. To this end I annoyed the missus greatly and bought 11 used Siemens 2x8mm feeders. Unfortunately they took a lot to get going (used really means used) and in the end only 9 work but that's still 18 reels of automated goodness.

After a burst over Xmas/New years I've made good progress. Not quite there but almost close to a functional machine..

Just a single fixed head to get underway but have done firmware / solenoids / pneumatic plumbing for 4 separate heads with pretty much the same pneumatics as used on my first machine. Still not sure if I'll rotate the head or the PCB, with my (non) engineering ability PCB is way easier and keeps the mass off the gantry. But if I want to be compatible with SCSi's software (and I'd like to) I expect I'll be building a rotating head at some point.

Main custom interface PCB is a PIC 18F4xxx as before getting commands from Mach3 via the parallel port

I've done the PIC firmware & all the other hardware interface PCB's etc are fitted / connected / working from the PIC board & including the board to drive the Siemens feeders.

Haven't touched the PC side yet though (Excel generating GCode, also Mach3 macros for controlling PIC interface functions)

On the mechanical side just still to do the PCB holder & non automated 'feeders' at the other end. as well as find a home for the upwards camera (solely for getting head offsets) & the bump pits for the component alignment.

Additionally I've some 0.9 degree steppers ordered to help with accuracy (I hope).

The machine is almost together for some functional/accuracy testing and then I'll start on the PC software.

Hopefully its accurate enough, there's a few things to change later but many areas are 'close enough', in fact, I learn't a lot on this one (my first CNC mechanical type build), and already a quite few things I'd change if I ever got to a MK3 version..

Finally, I really must learn to use some sort of 3D CAD package one day...