Pick & Place migraines

[cyberhedz]

Hello everyone.  I've spent the last 4 weeks trying to find the perfect PnP for my projects and I've come around full circle and still don't have a choice I'm happy with, mostly because the model I picked, has no reviews.  I originally looked at the neoden TM series, then I moved on to Madell, manncorp, smtmax, etc.  I'm looking for an entry level machine, under 10K USD, with vision and BGA/QFP support, and preferably 0201 capabilities.  Can we have a small discussion on these and perhaps get a review or two from the wild on teh Madell PnP's?  I was particularly interested in the Mini-X1 (http://www.ntscope.com/Mini-X1.html).  Perhaps I overlooked some other brand?

Must haves:
BGA/QFP support
computer vision
desktop size
under 10K USD
support options


niceties:
0201
integrated pump
comes with reel holders


Thanks everyone!

Sam

[mazurov]

You won't find a perfect PNP, ever. Figure out how much money you're willing to spend and then pick the machine which offers the most features for that amount. I would prefer more reels over vision; besides, for small pitch parts you'll need _bottom_ vision - I don't think you will find one for the money you've mentioned, especially if you're shopping for desktop models.

I'm very happy with TM240A. It works well, has enough reels, and tech.support is very good. It is also very simple mechanically; I'm slowly building an upgrade kit to switch control to openpnp, add top/bottom vision, more reels, plus a vibratory feeder or two.

[cyberhedz]

This is exactly the kind of response I was hoping for.  The mini-x1 does have bottom cameras, but your "upgrade kit" intrigues me.  How far along are you?  I'd gladly test with you and grab a TM240, but the vision is a must as I plan on placing an Altera Cyclone FPGA, which is only available as a TQFP or BGA.  Honestly the Mini-X is exactly what I'm after, but I haven't heard anything good or bad about this model, and only saw a single post from 2012 saying their support was a joke and it was a waste of money.  For $6,700, I want to know I wont be getting burned.


thanks for the feedback

[mazurov]

What is the problem you're trying to solve? If you need to place small pitch parts consistently buy yourself a microscope - you'll get better results, spend far less money and you will need a microscope anyway when you finally get your machine, with or without vision, and start writing and testing job files.

A machine will make sense if you spend more than 20% of your waking hours placing parts. You still won't be able to spend less time but you will be able to stuff more boards. Vision starts making sense when your volumes are high, otherwise simply inspect every panel after placement under a microscope and fix placements that you don't like. BTW, my machine places 0.5mm pitch TQFP spot on from the front tray; I still inspect every panel before soldering.

[cyberhedz]

My problem I'm trying to solve is that I have VERY shaky hands (probably from the caffeine i drink).  I'm wanting to place components on roughly 250 boards for now, and the continued production of other boards.  The less i can spend the better, but I need to find a solution that will work.  Thank you for the microscope suggestion, that would be about half the cost if I bought that with a TM240, when compared with the Madell. 

[mazurov]

My problem I'm trying to solve is that I have VERY shaky hands (probably from the caffeine i drink).

With TM you'll need manual dexterity too for certain tasks (and I heard Madell is harder to operate than TM). Neoden have instruction videos online -> https://www.youtube.com/user/NeoDentech/videos , you may want to watch them to see if you can handle the machine.

[cyberhedz]

those videos dont make it look too hard, but according to the sales rep at madell, I just send the files over straight from Eagle, and it will use vision to calibrate and tune, then place the parts. "full automatic" is what the rep said.  Liek I said though, he was a SALES rep, so he could be blowing smoke just so i buy from him.

[mazurov]

Sales ppl are socially accepted crooks nowadays, don't pay attention to them unless you enjoy being taken advantage of. Personally, when I was choosing my machine I have never seen/heard a single good word about Madell from anybody, and only horror stories from owners. Also, if you're a newbie your first priority must be a machine that is popular, for obvious reason.

[cyberhedz]

fair enough point.  I'm taking from the looks of the boards here, the TM240 is VERY popular, so I'll put it back at the top of my list.  Well, second place, first place is still in the air (local company MIGHT be selling their smaller machines)

[coppice]

fair enough point.  I'm taking from the looks of the boards here, the TM240 is VERY popular, so I'll put it back at the top of my list.  Well, second place, first place is still in the air (local company MIGHT be selling their smaller machines)

How do you judge its popularity? Its talked about a lot, because its the cheapest thing around. Does that actually translate to real users? I see lots of comments, but I don't see many real world users telling their tales of success or failure. mazurov seems pretty positive about it, but does his use case match yours? You need to hear from a few real users before a picture starts to form.

[mazurov]

fair enough point.  I'm taking from the looks of the boards here, the TM240 is VERY popular, so I'll put it back at the top of my list.  Well, second place, first place is still in the air (local company MIGHT be selling their smaller machines)

How do you judge its popularity? Its talked about a lot, because its the cheapest thing around. Does that actually translate to real users? I see lots of comments, but I don't see many real world users telling their tales of success or failure. mazurov seems pretty positive about it, but does his use case match yours? You need to hear from a few real users before a picture starts to form.

Here's the TM2[2,4]0A thread at DP -> http://dangerousprototypes.com/forum/viewtopic.php?f=68&t=4903&sid=12867583eab38792ade2ae326a98c973 , with plenty of people with machines telling their tales and posting pictures.
 

[cyberhedz]

That's the thread I found during my first wave of research.  Still trying to find the right machine for me, it's a hard one, it's easier to buy a house.....

[odingalt]

I was in a similar boat a few years ago.  I ended up getting a hand place machine.  The fact was that, at least back a couple years ago, there just weren't any machines capable fo placing 0402 and sub 0.50mm pitch component on a $10k budget.  If you only need 0603 and 0.65 pitch then you can place them by hand with a manual assist placer and save a lot of fixed cost.

http://www.abacom-tech.com/Manual-SMT-Pick-and-Place-Machine-ezPick-P93496.aspx

The Chinese also sell a hand assist similar to the MannCorp except it comes with a camera system I think it was around $2k-$3k (basically this model with a camera:  http://www.manncorp.com/smt/prod-179/smt-place-2000-manual-pick-and-place.html)

When you hit about a $20k budget then your options open up greatly on the used/second hand market.  It's a buyer's wet dream right on now second hand equipment but the window is closing as the economy could turn around starting as soon as this upcoming x-mas.

[cyberhedz]

hmm, I scoffed at first, but the more i think about it, it might just work.  The Ez-pick, granted is FULLY manual, but it could be the assist i need to knock out these 250 boards, start generating revenue so I can open up my options to the 20K PnP machines.

I'm still considering the TM for now, just for the hackery of getting to screw around and maybe get vision added to one.

[coppice]

Here's the TM2[2,4]0A thread at DP -> http://dangerousprototypes.com/forum/viewtopic.php?f=68&t=4903&sid=12867583eab38792ade2ae326a98c973 , with plenty of people with machines telling their tales and posting pictures.

If you scan through that list most of the user comments come from a very small number of people, and its hard to draw much of a conclusion. E.g. one person is stuffing large panels every day, while another has alignment problems when the gantry works over the whole board area. Quite a few of the overall strengths and limitation do come out in those exchanges, which are not obvious from the blurb - e.g. parts must be <3mm high - but its the overall predictability of operation with it working day in and day out that people really want to flush out of the discussion. It seems several of the heavy posters have ultimately put their TM240A aside, and it is not entirely clear why.

[Kjelt]

IMO 250 boards is way over the line to place manually, depending on how many components have to be placed per board ofcourse.
If you only calculate your own labor costs for a reasonable hour fee you will spent probably more then if you let a contractor do the pasting, placing and reflowing for you. They often also do a fast automated flying probe QC so that is also time saved. Owning an affordable PnP machine made in China might sound great till you get issues and need repair. Then China is far far away from helping you.

[mikeselectricstuff]

hmm, I scoffed at first, but the more i think about it, it might just work.  The Ez-pick, granted is FULLY manual, but it could be the assist i need to knock out these 250 boards, start generating revenue so I can open up my options to the 20K PnP machines.

If the issue is to get things moving, don't waste time & money on kit that will be standing idle most of the time - subcontract it out.
Unless you need very fast turnround, doing runs of hundreds of boards in-house makes little sense

[odingalt]

IMO 250 boards is way over the line to place manually, depending on how many components have to be placed per board ofcourse.
If you only calculate your own labor costs for a reasonable hour fee you will spent probably more then if you let a contractor do the pasting, placing and reflowing for you. They often also do a fast automated flying probe QC so that is also time saved. Owning an affordable PnP machine made in China might sound great till you get issues and need repair. Then China is far far away from helping you.

I placed over 25,000 components manually by tweezer before I purchased the manual picker arm (for me specifically the ABA com EZPick and I highly recommend it).  Then I placed another 50,000 by hand with the picker.  Here is a video of just one of thousands of panels I processed by tweezer or manual picker arm.  That manual picker saved my life and is actually relatively quick.



If you are a cheapskate (like me) and absolutely need to save on cash, then you price your labor at $0 and get it done for "Free".  Of course after 5 years of doing this kind of work for profit, I am burned out in general, and I no longer price my own labor at $0 when costing jobs for customers.  And we were lucky enough we were able to grow business enough to budget about $80,000 for a good quality vision pick and place, conveyor reflow, etc.

Everybody's gotta start somewhere, money sent to a contractor is money that doesn't go back into your own business.  If you're in it for profit and for the long haul you must reinvest as much in yourself as possible if you want to grow, it's a very competitive world out there.

[nctnico]

What is the problem you're trying to solve? If you need to place small pitch parts consistently buy yourself a microscope - you'll get better results, spend far less money and you will need a microscope anyway when you finally get your machine, with or without vision, and start writing and testing job files.

Wouldn't it make more sense to just outsource production? Spending 10k (or more) on a machine justifies a lot of time spend by someone who specialises in assembling boards. It does take some time & effort to find a cheap assembly house though but I'm sure people on this forum can help with that. Your time is probably better spend doing design work or at a daytime job.

[cyberhedz]

Loving the feedback here, thank you everyone.  I think I've decided to build my own Cartesian style manual placer, similar to the EZ-pick, along with a microscope.  This should get me through my first thousand boards or so, enough to start generating the profits needed to source a larger, supported, reliable, PnP.

using this design as my starting reference:
http://vpapanik.blogspot.gr/2012/11/low-budget-manual-pick-place.html

[nctnico]

Loving the feedback here, thank you everyone.  I think I've decided to build my own Cartesian style manual placer, similar to the EZ-pick,
http://vpapanik.blogspot.gr/2012/11/low-budget-manual-pick-place.html

Hmm.. That looks like an interesting device. I do see some problems though. Components (especially resistors) stored in bins may need to be flipped over before placing.

[Kjelt]

Yeah don't use small components that you use in huge quantities in bins, use a piece of the reel strip, resistors are cheap $6 for a reel of 5000. Choose 20 most common values and use them as much as possible in your designs. The few resistor values occasionally left can be flipped by hand. Look at the video of mikes electric stuff for some handy tricks for manyal p&p.
Btw if your hands really shake it might be a good idea to have some form of brakes on the x,y axis guidance so that you only have the z axis to worry about

[cyberhedz]

I should be able to use this, my shakes are usually just enough to make SMD a pain, not a parkinsons shake.  the brake could work, but I think this will be sufficient for long enough to just say screw it and build my own semi or even full auto PnP

[JohnSL]

I had the need to make about 100 circuit boards and didn't want to do them by hand--I really hated trying to get LEDs out of the tape and oriented correctly. So I designed and built a manual pick and place machine:

http://trains.socha.com/2014/07/building-manual-pick-and-place-machine.html

The cost was about $50, and it works great. It's faster and more precise than doing it by hand.

[coppice]

I had the need to make about 100 circuit boards and didn't want to do them by hand--I really hated trying to get LEDs out of the tape and oriented correctly. So I designed and built a manual pick and place machine:

http://trains.socha.com/2014/07/building-manual-pick-and-place-machine.html

The cost was about $50, and it works great. It's faster and more precise than doing it by hand.

It seems like adding solder paste dispensing with your gadget, by integrating one of the inexpensive compressed air dispensers (e.g. the 983A you can find on TaoBao or E-Bay), might make a pretty useful prototyping tool. Those dispensers are less than $50, although you need an additional compressor.