need prototype parts made now, and small production batches later

[bootstrap]

I designed what might be called a very simple "manual pick-and-place machine" for placing very tiny and very fine pitch electronic components on PCBs (printed circuit boards). I need someone (or some two or three) to:

#1: help me make a couple final design decisions.
#2: machine parts for one or two prototype devices.
#3: machine small volumes of parts thereafter assuming it works well.

project nature
I designed a tiny high-speed, high-resolution video camera for real-time robotics applications, then designed PCBs to build prototypes, and had those PCB made. Now I needed to assemble those PCBs so I can develop the FPGA firmware and microcontroller software to finish this device. But some of the components are just too small or have too fine pitch contacts to assemble manually. The usual approach is to buy a $50,000 pick-and-place machine, then spend another $10,000 or so on solder reflow ovens. That's more than many individuals and startups can afford to spend, which has greatly reduced the flow of cool new electronic gizmos from individual engineers and startups.

I figured out how to create a very simple, inexpensive device that can place small and small-pitch components to an accuracy of 0.001" ~ 0.003", which is good enough for this purpose. Now I need to create the prototype to populate my own prototype PCBs so I can finish my robotics camera project.

But this device should also be attractive to a great many engineers and startups to help them develop products and grow to the point they can afford conventional automatic pick-and-place machines. I'm confident these devices can be manufactured for under $500 and sold for under $1000, which is a far cry from $50,000 (the machines cost about $35,000 but you need all sorts of component reels and accessories to make them practical, hence the $50,000 that I mention).

I'm not interested in making lots of money making and selling these devices, but do consider this a modest opportunity for myself and someone with modest CNC machinging capability or better. So after the prototypes "work great", I'd like someone to manufacture parts for what I assume will be modest production runs (my estimate: 100 units every 1~6 months).

The "design help", "1~2 prototypes" and "production" can come from one, two, three or more people. If you can potentially help me in one or more ways, please contact me for detailed discussion.

product description
The device is essentially a simple x-axis stage that slides left and right a few inches above the benchtop. On the x-axis stage are two holes. Into one hole is attached a cheap "digital microscope", which is essentially just 5 megapixel webcam circuity built into a ~25mm diameter tube by ~100mm long tube with a lens that focuses on objects very close to it (25mm~50mm). Into the other hole is attached a z-axis stage to raise and lower a ~1mm diameter vacuum pickup tip to pick up electronic components.

That's all, except a simple 4 legged stand that sits on the workbench below this device. The stand holds a PCB about 25mm above the table-top, which is enough to assure most components already mounted on its bottom side do not touch the table. The front rails of the stand is about ~60mm wide and has one or more 50mm square cutouts to hold "component templates"... one template for each component pitch. Each "component template" is a ~50mm square piece of thin metal (2mm~3mm) with an x,y grid of holes of a specific size and pitch so the balls of a BGA package of the corresponding pitch rest in the holes, and align the balls on the BGA part with the "component template" holes (the "component template" holes are only about 1/2 to 3/4 as large as the BGA balls that rest in them).

That's everything, except I will need to develop a software application that performs the appropriate steps necessary to assure components are placed correctly. But I'm a software developer, so I can do that.

project operation
To place a component, the following steps are taken:

#1: The operator slides the x-axis stage to its right-most limit, and secures the x-axis stage in place so it will not move.

#2: The operator inserts a "component template" with the same pitch as the BGA package he wants to place into the ~50mm diameter square opening on the PCB holder. The operator then slides the PCB holder until he sees the "component template" appear on his computer LCD screen in the application I write. At this stage, the software application simply displays whatever the digital microscope is pointed at, but also displays thin vertical and horizontal lines that intersect in the center of the window on his LCD screen.

#3: The operator slides the PCB holder around until the center hole or mark in the "component template" is centered on the crosshairs, and the rows and columns of holes are also centered on the horizontal and vertical lines of the crosshairs. The "component template" is now exactly aligned with digital microscope and its crosshairs. The operator then presses the space-bar (or clicks a button on a GUI) to indicate "component template centered and aligned". The software application captures the image at that moment, then henceforth displays both the image of the captured image and the continuously updated image of whatever is below the digital microscope.

#4: The operator now sets the electronic component on the "component template", and feels the component fall or detent into the "component template" holes. The contacts of the component are now exactly aligned with the digital microscope and its crosshairs.

#5: The operator now unlocks and slides the x-axis stage from its right-most position to its left-most position, which is exactly 50mm travel. 50mm is also exactly the spacing between the optical axis of the digital microscope, and the axis of the vacuum pickup tool. The operator then secures the x-axis stage in place at its left-most position so it will not move. This places the tip of the vacuum pickup tool exactly over the center of the component.

#6: The operator lowers the z-axis stage until the tip of the vacuum pickup tool presses lightly on the component. The tip of the vacuum pickup tool is lightly spring loaded downward, so when the tip contacts the component, and the downward motion continues 1mm ~ 2mm further, the tip stays at the same position on the component. The extra 1mm ~ 2mm of downward motion lightly compresses the spring, triggers a tiny pushbutton or limit-switch, which generates a tick or tone sound and opens the vacuum valve which feeds through the vacuum pick-up tip and holds the component to the vacuum pick-up tip.

#7: The operator hears the "contact" indicator tick or tone, and then raises the z-axis pickup stage to its upmost position (about 25mm~35mm motion).

#8: The operator slides the x-axis stage to its right-most limit, and secures the x-axis stage in place so it will not move.

#9: The operator watches the window on his LCD screen as he slides the PCB holder around on the benchtop. When the contacts for the desired component appear in the window, the operator then moves and rotates the PCB holder until the continuously updated image of those contacts are exactly aligned with the image of the "component holder" captured in step #3 above. At this time the component contacts on the PCB are exactly in the same position and alignment with the x-axis stage and digital microscope as the holes in the "component template" were in step #3.

#10: The operator slides the x-axis stage to its left-most limit, and secures the x-axis stage in place so it will not move. This places the component contacts or solder balls on the bottom of the component exactly over the contacts on the PCB.

#11: The operator lowers the z-axis stage until the component is placed on the PCB, then during the next 1mm ~ 2mm downward motion the vacuum tip is not moving as the spring is compressed... until the tiny pushbutton or limit-switch is triggered, which releases the vacuum (and possibly generates a very slight positive puff of pressure to make sure the component drops off), and generates a click or tone to tell the operator the component is placed and the z-axis stage should be raised.

#12: The operator raises the z-axis stage.

The component is now precisely placed.

The "component template" for QFN packages are simply square holes the same size as the OD of the QFN package. Otherwise the process is identical to the description above for BGA components, except the crosshairs might be slightly different (4 crosshairs separated by the appropriate distances so they exactly bound the QFN package).

product design
The x-axis of my current design is based upon four 20mm ID linear bearings in pillow blocks sliding on precision 20mm OD hardened steel shafts.

The z-axis of my current design is based upon two 20mm ID linear bearings pressed into a machined aluminum tube with a boss on one end that bolts to the x-axis stage.

The z-axis itself is hollow so we can connect a vacuum tube to the top and provide suction through the vacuum pickup tool at the bottom. Unfortunately we need another ~12mm OD with concentric and hollow z-axis shaft within the ID of the main z-axis shaft so we can tolerate the operator lowering the z-axis stage slightly further than necessary and compress the spring and trigger the tiny pushbutton or limit switch to toggle the vacuum off and on (presumably through some cheap and standard vacuum toggle device that I have not yet found). I have a feeling that some alternate configuration is possible, simpler and better than having the two-level nested concentric z-axis shafts... but I can't quite visualize it. How about you?

product components
#1: (2) end-pieces : sit on workbench at far left and right and hold the two 20mm OD steel shafts.

#2: (1) x-axis stage : a flat piece of metal with two 16mm~25mm holes to mount the digital microscope and z-axis stage/shaft, plus several other threaded and/or countersunk holes for attaching these and other components to the x-axis stage.

#3: (2+2) limits : Presumably we press or shrink/expand fit (2) hardened stainless steel balls into the left and right ends of the x-axis stage to hit against the ends of (2) hardened steel pins in the end-pieces to establish the "left-most" and "right-most" limits of the x-axis stage.

#4: (2) x-axes : 20mm hardened [stainless] steel shafts.

#5: (4) x-axis linear ball bearings : 20mm ID linear ball bearings in pillow blocks that bolt to the x-axis stage (#2).

#6: (1) digital microscope : 5-megapixel digital microscope.

#7: (1) outer z-axis : 20mm OD x 12mm~16mm ID hardened [stainless] steel shaft, probably about 100mm long

#8: (1) inner z-axis : ~8mm OD x ~2mm ID hardened [stainless] steel shaft, probably about 50mm long.

#9: (1) z-axis housing : machined aluminum tube with retaining clips.

#10: (2) z-axis bearings : 20mm ID linear ball bearings.

#11: (2) z-axis bearings : ~8mm ID linear ball bearings.

#12: (1) vacuum pickup tip (need to design).

#13: (1) vacuum pickup tip spring (need to design).

#14: (1) vacuum pickup tip limit-switch (need to design).

#15: (1) vacuum pump (cheapo aquarium pump or something simple).

#16: (4) PCB holder legs : 20mm square extruded aluminum bars?

#17: (?) PCB holder frame (need to design).

#18: (8) component templates : each 50mm square x ~2mm thick with holes.

#19: Did I miss anything?

request for help
#1: Help to finish and/or modify design (to improve or make easier to fabricate).

#2: Fabricate parts for 1 or 2 prototypes.

#3: Fabricate parts for batches of 25 to 100 once development is complete.

-----

Thanks for reading this long, detailed post. I welcome any constructive ideas and suggestions and offers or proposals to help. I look forward to hearing from a few of you. I'm temporarily located in the southwest USA for the next few months, but I don't see why we need to be nearby, or even on the same continent necessarily.

Thanks.

Max

[IJF]

PM sent. Thank you.

[Tormachmaster]

I may be interested over 22 yrs cnc have my own prototype shop you can either send a picture and approx dimensions or a model for a quote I dont think i want the entire project will decide after seeing your parts but have resources of other shops that may help or you may find a few good guy's here on the zone.

[rolfsnes]

I I am interrested in making these items.
Have a CNC workshop on 1000sqm with different machines.
Please send a quotation to post@bechermaskinering.com
with drawings and materialtype , of what you want to have made.

[bostosh]

Interesting project,
The expertise to do this is here on the Zone.

Do you have a complete 3-D model or a concept sketch?
This will drive the machined parts requirements.

The verbal description is good,
The expected tolerances are do-able.

This can be done, think of those who assemble those Swiss watches one at a time. Or the huge complex injection molds.

[Dualkit]

Originally Posted by bostosh Interesting project, The expertise to do this is here on the Zone. Do you have a complete 3-D model or a concept sketch? This will drive the machined parts requirements. The verbal description is good, The expected tolerances are do-able. This can be done, think of those who assemble those Swiss watches one at a time. Or the huge complex injection molds.

True, but there are also hacks that will destroy your confidence in machinists. I have been sent some samples of someone else's work along with a stack of blueprints recently on a product a customer wants to get to market. Poor customer had a product he was close to perfecting and someone made him a mess of garbage that he couldn't even work with and he paid upfront. He might have found the two hacks that did the work on the sister site. My advice to Bootstrap, have a phone conversation with a few guys, maybe ask for references. Beware of people who brag on themselves too much, most good machinists let their work do the talking. Size of the shop and age of machinery is not as important as the guy or gal doing the work and their attention to the job at hand. I know guys that couldn't run a nice part if they had $1,000,000 of equipment and a 50,000 square foot shop, and I have seen guys do heart surgery grade work on nothing but WWII era manual machines.

[bootstrap]

Thanks to everyone for your helpful comments. I talked with two people and exchanged emails with two others, which has helped much with making a few design choices.

I am working on some drawings in CAD software now and when they are finished, I will forward them to everyone for comments and possible quotes. This may take a little while longer than normal because I have never worked with a mechanical CAD program before (only PCB design software), so I'm trying to learn the CAD software at the same time as I design the remaining parts. Time to step up from the old ruler, paper and pencil it seems.

I will send you output files and/or PDF drawings when I'm done, or sufficiently close to done to get feedback.

Regards,
Max