Homemade Pick And Place Awesomeness

[Sionyn]

[mikeselectricstuff]

Yawn....
Too slow to be of any use.
When will people stop wasting time pissing about with X/Y gantries and realise that the problem that needs solving is feeders?

[David_AVD]

Sure looks slow.   

[EEVblog]

A noble effort of course, but I agree with Mike.
There is no practical point doing a DIY PnP machine unless you get the speed, and the feeders right.

Dave.

[Psi]

Whats the issue with DIY feeders being hard? 

Surly it's just a motor and some sprockets to pull the tape plus a few tensioners to keep it tight?.
or am i missing something critical?

[mikeselectricstuff]

Whats the issue with DIY feeders being hard? 

Surly it's just a motor and some sprockets to pull the tape plus a few tensioners to keep it tight?.
or am i missing something critical?

It's far less trivial than it first appears.
You need to keep the tape flat, and have a linear advance ( a large sprocket wheel won't work for a few reasons), peel off the cover tape with the right amount of tension, cope with paper and plastic tapes without sticking or jamming, avoid any sudden motion that can bounce parts out of paper tapes, avoid anything that needs oil to lubricate as it will gum up with dust from paper tapes, while keeping the unit cost low and the width small enough to allow room for plenty of feeders. You also need to make versions for different widths. Although larger sizes (16mm and up) could reasonably be accommodated with an adjustable width lane as there probably won't be many, making smaller ones adjustable would waste space.
Bonus points for minimising the amount of leader and trailer length required to minimise part wastage.

I'm not saying it can't be done, its just that most DIYers have failed to realise that this is the problem to solve before you even look at the actual picking and placing.

One solution to the width problem is multi-lane feeders, which can also work out cheaper per-lane, however these come with  their own problems, in particular flexibility in lane size mix, and inability to swap or remove indvidual lanes for easier loading.   

I think the best avenue for cheap DIY feeders may be to throw cheap computing at it - clever vision of loose parts in small trays (e.g. OpenCV) with a vibrator motor to shake them up.

[tom66]

At my former workplace we had a 20 year old pick and place (cost only about £5k) but it had some lightning speed over this, at least 10x faster, and it was vision based, ran Windows 95 I think. The feeders were £180 each and it had some 30 odd -- these were pretty intricate pieces of equipment. I wonder if the "solution", for now, is to get hold of some scrap/second hand feeders and mod them to work with the unit?

[mikeselectricstuff]

At my former workplace we had a 20 year old pick and place (cost only about £5k) but it had some lightning speed over this, at least 10x faster, and it was vision based, ran Windows 95 I think. The feeders were £180 each and it had some 30 odd -- these were pretty intricate pieces of equipment. I wonder if the "solution", for now, is to get hold of some scrap/second hand feeders and mod them to work with the unit?

Yes, but the problem is that feeders for old machines that are still out there are likely to be in more demand than the machines themselves. May be a possible if there is a machine out there that suffers from some common catastrophic fatal failure mode that renders it worthless.

[JuKu]

I agree that any placer needs to be faster and more accurate than manual method. Also, for a production environment, the feeders are the problem, and there is no good solution: You need a few ten of those, and even at a couple of hundred each, the investment becomes big. I haven't had to need to do the math, but I suspect that the feeder cost fast goes so high, that it makes sense to buy a real commercial machine, with support, warranty and all that.

However, I don't think the feeders are an issue in a DIY or prototype lab. In that environment, you typically would be building one or two boards at a time. Throughput is not an issue. What the machine needs is a big enough work area and tape holders. I can see a system where you would have many slots where the machine can pick components along the whole slot. You would peel the cover tape off manually, load the slots and tell the machine that "C1-32 (100n, 0805) are in slot 1, positions 1-32", "R125-130 (2k2, 0603) are in slot 5, positions 5-10" and so on. Add to this the capability of picking up loose components (with help of down-looking camera and clever software) and an up-looking camera for BGAs. I think a simple DIY kit/prototype level machine is viable and has its place in the world. None exists yet, though.

But it still needs to be fast and accurate enough.

[David_AVD]

Sticking the tapes down and hand placing them with a vacuum tool is actually pretty quick.  Much faster than the machine in that video.  I think Mike showed SMD hand placement in a video a while back?

[EEVblog]

However, I don't think the feeders are an issue in a DIY or prototype lab. In that environment, you typically would be building one or two boards at a time.

In that case it's pointless having a pick'n'place machine IMO.
As the setup and fiddle time would not be worth it for only a couple of boards.
You'd be much better off financially and time-wise getting a simple vacuum hand placement unit.

Dave.

[David_AVD]

Even doing 100 boards is not out of the question with hand placement.  Where you'd save more time is with a paste stencil imo.

[mikeselectricstuff]

However, I don't think the feeders are an issue in a DIY or prototype lab. In that environment, you typically would be building one or two boards at a time.
Throughput is not an issue.

The problem is that if you're only building a couple of boards, the setup & feeder loading time means that any P&P machine, however fast, just can't be justified. 
I have a pretty streamlined process to set up jobs on my P&P from the PCB CAD data, but for small jobs I still hand place as it's quicker. 

A lot of people haven't realised that there is simply no market or use for a P&P machine that's significantly slower than hand placing. 

I'm sure at some point, some idiot will try a kickstarter for a P&P design, but if so it will either not meet target, or will end in tears.

[mikeselectricstuff]

I think Mike showed SMD hand placement in a video a while back?

[jucole]

I think Mike showed SMD hand placement in a video a while back?

totally Superb!!  I've never seen a utube with so many great tips-per-second as that one!!  many thanks Mike!

[JuKu]

However, I don't think the feeders are an issue in a DIY or prototype lab. In that environment, you typically would be building one or two boards at a time.

In that case it's pointless having a pick'n'place machine IMO.
As the setup and fiddle time would not be worth it for only a couple of boards.
You'd be much better off financially and time-wise getting a simple vacuum hand placement unit.

Dave.

Maybe, maybe not. For a size of board Mike shows in his video, I totally agree. On the other hand, there are a few cases for a machine, even in a small lab:

This is the board I did last time (an 8 channel DAC):

Doesn't look that complicated (and really, isn't), and it definitely is in a hobbyist or a small development lab range. Still, there are 450 SMD components on that board. Doable by hand, but somewhat on the heroics side. The paste started to turn bad before it was done (the board turned out all right, though), and my hand started to be shaky. It is slower than you would think when the board has variety of components and you have to look up each one. Even a system that would point out the location of each component in turn would have much helped. And yes, i have a manual vacuum unit with a foot-operated component release. I'd pay for a machine that would give significant help on this before I do another similar board. There is none out there, though.

The second point is BGAs. Soldering them is not an issue: I was pleasantly surprised of the profile accuracy and results from my very cheap pizza oven method. The oven/controller combo is capable of accurately following the manufacturer's recommended soldering profile, so I can't see no reason why my oven wouldn't be good enough for BGAs, even though it is about the cheapest there is - if I just could place them.

Small and fine pitch components are also an issue, although not that much since you can see what you are doing. Still, for us older guys 0.5mm pitch ICs and 0603 and smaller caps/resistors are a challenge, no matter how good vacuum tool or microscope. I have no complaints at my tools, but I don't think I could do more than 20-30 0603 resistors in a day; I know I can't do any 0402 by hand.

[mikeselectricstuff]

Even a system that would point out the location of each component in turn would have much helped.

Wouldn't be too hard to do, maybe a projector?
In the old through-hole days, you could get a system that used a film strip to project a dot onto a board showing the location, and had a series of part bins, each with a light that would light up to show which part to pick. There was also a version with columns of bin carousels, so a bin would rotate into view as required.
However just like a modern P&P, it needs a lot of setting up.

And yes, i have a manual vacuum unit with a foot-operated component release.

Unless you have very few different part values, setup time and loading feeders would probably take a similar time to hand placing a board like this

I'd pay for a machine that would give significant help on this before I do another similar board.

How much? And how much setup time would you be willing to spend?

There is none out there, though.

There is. it's called a proper pick & place machine.

[Otatiaro]

Hello,

Did it too ...



Takes a few hours to solder about 10 boards, including inspection and programming.

If (when ?) I had the money, I'd buy an MDC co Luna Vision pick and place, the one that can handle cut tapes (with a few feeders too, for common parts - 10k res, 0.1uF cap, etc - just in case).

Total cost is between 30 and 40k$ ... not that bad if I BRAINs start to sell a bit faster (http://www.msheli.com/Item.aspx?IDItem=330).

If you think about it the professional way, takes one day to solder 10 boards, one day of work worth 500€, so once you prepared 600 boards with it then it's self paid (say 1000 boards just to make sure).

Thomas.