PCB Production

Hullo guys

I don't know if anyone has thought along these lines (Probably, and being newbie to this group I am probably going over old ground, so sorry in advance).

As an open source junky I think open sourcing machinery to build better machinery is the muts nuts.

I have been reading a number of articles re RP and PCB production using some wonderfully novel techniques. If a little high tec.

PCB production from standard packages is usually anything but trivial. The higher the required quality and layers the messier and less eco friendly the process is.

In the spirit or reprap how about machines that can take part in the production of their own next gen PCB's using some simpler more established kitchen table technology .........

Low Tec

Has anyone though of using a syringe tool head or maybe even drawing pen tool head (disposable permanent ink type perhaps with either spring or gravity providing correct pen pressure) to dispense water/etchant proof tracks onto copper clad board before direct etching ????

Higher Tec & more messy but potential finer resolution (next reprap gen maybe).

UV Light source (shutter controlled) with a cheap fiber optic connected pen to draw on photo etch board, then develop, then etch.

Using tooling as is.

I guess even using the extruder to draw tracks as the plastic will be etch proof. Then etching the exposed copper.

Providing the start points on the boards can be synchronized front to back (ie the registration and repeatability is accurate enough) and folk don't mind soldered via's double sided boards as well as single sided should be possible using these methods.

Withs sufficiently fine drawing and control single sided boards for surface mount components should be a breeze (no holes to worry about)

Thoughts for what they are worth....

aka47

--
---------------------------------------------------
If sense were common:-
1. Everyone would have it.
2. The exploited wouldn't be.
3. The exploiters would have to get their sleeves
rolled up and make a useful contribution like
everyone else. (or preferably starve). 4.5i

hey there,

we have definitely considered doing PCB production... if we are to truly self-replicate one day, then PCB's are something that we will need to figure out how to create.

if you are interested in this area, i highly suggest getting a machine and starting to tinker with it. the sooner we have people actually researching it, the sooner we'll have the technology fully developed. remember: failures are almost as valuable as successes from a learning standpoint.

that being said, this is probably a 2nd generation type technology. right now we're focused on getting a solid cartesian bot base to work from, as well as nailing the electronics and software. once we have those fundamentals thoroughly worked out and tested, then it provides a platform to do all sorts of experimentation: pcb creation, laser stuff, etc.

2008 is going to be very exciting =)

Funnily enough after I had posted the above I found your page on the wiki which seemed to suggest pretty much what I had posted. Sorry

Yup I agree getting a machine up and running even if limited is a good place to start.

I pretty much have all of the PIC development stuff plus a bunch of electronic test gear (It is what I do) being out of work at the moment though is somewhat limiting my expenditure beyond the reasonable.

I can offer brain work and help with coding etc but at the moment acquisition of physical reality is tempered by fiscal reality.

I think updating the repstrap strip board stuff and coming up with an ultralowcost strap project to take folk to the next level including PCBs is probably a worthwhile effort.

What would the minimum dimension (XYZ) of workspace to be able to fabricate a reprap particularly including PCB's ??

aka47

ps the afghan lathe is pretty cool

i'm not entirely sure what the largest size will be. i know that most of our circuit boards are no wider/longer than 3" however, they will probably have to be redesigned with lower tolerance design rules. right now we can make them pretty small since the commercial places have very accurate processes.

also, once i feel happy with the status of the next-gen boards, i'm going to put in some production quantity orders of the PCBs which should hopefully drop the price by 50%. i'll make a post when that happens.

The fab&home folk seem to have got a workable solution for the syringe dispenser. if coupled with needles (cut off and deburred) used as fine nibs pretty fine dispensing of etch resist ink should be possible.

Not to mention solder paste.

A syringe fitted with a rubber tip and operated in reverse is potentialy a vacuum pick and place tool for surface mounted components.

Just going off to see if I can find an online lab ware place that can tell me the standard sizes etc for syringes and compatible needles.

aka47

i agree... for this type of thing, a syringe extruder seems like a great candidate. the awesome thing about syringes is that they are ridiculously cheap thanks to the medical community and having to use disposable syringes.

i dont have the name of the company online, i'll look it up at the lab tonight, but you can get some big syringes for $0.50 each, and needles for even cheaper. google medical supply, or needles, or something like that.

~Zach

Been mooching around t'internet.

Hypodermic tubing the stuff the needles is made from and therefore the needles seems to be sized in standard wire gauge (SWG or Gauge).

What we realy need is a sensible conversion chart for a given manufacturers needle sizes (ie nominal Internal Diameter).

Some folk appear to turn out syringe needles that are blunt too. (Pharmacy filler needles) which would'nt need cutting and deburring.

Just need a sensibly small internal diameter or a range of them to try with some etch resist ink. (Viscosity issues ????).

Etch resist ink looks to either come in-pen or in huge quantities for screen printing (Would this be too visouse the manufacturers do quote shore values but they are a touch meaningless to me, anyone help here ??)

May have found something sensible though, need to look into it.

A last thought for the night, if there is'nt sensibly priced etch resist ink of a low enough viscosity for the needle gauge we want to use. Could we use a more usual drawing ink as a photo mask over presensitised board.... and maybe a standard rotoring style drawing pen.

Must admit to fancying the syringe thing thoug as it would have other uses too.

Off to bed

aka47

How about the nozzles used to inflate sports balls? They have a side port, so they'd have to be cut short, and I don't have one handy, so I don't know the internal, or for that matter, external, dimensions.

... i mostly order at Globaco ( [www.globaco.de] ), they have all needed components and needle-sizes - dunno if they ship overseas, so look if you have a similar supporter for dispensing- and soldering-robots locally.

My needles have a 'Luer-Lock'-adapter, what's pretty interchangeable as i use the same needles for paste-dispensing, vacuum-pick'n'placing of my micro-parts and for our air-micro-brushes for cleaning ...

Viktor

Viktor

I am in the UK so am technically in the Eu Zone. (Although many here seem to think we are attached to America, 5x state and all that) So purchasing/shipping should not be a problem

Unfortunately like many English speakers my German is savagely worse then the worst English any German uses (for German substitute most other eu languages). Yes its dire and inexcusable I know.

Is there any chance you could send me a link to the needles sizes table from globaco or educate me with a couple of German words that would allow me to fumble my way through their site.

Cheers

aka47

Hi aka47,

... here is the link to the needles (and syringes/cartoushes) [www.globaco.de] - go to the 3rd page (marked as page 8)

On the bottom of next page you have the colour-tables and hole-diameters in inches and millimeters.

They sell in bulks of 50 pcs.

Viktor

PS: when home, i'll try to find an english catalogue ...

Edit: --- and in the meantime you can watch a movie of a "comercial repstrap"
[www.aeb-robotics.com]

Edited 1 time(s). Last edit at 12/05/2007 03:59AM by Viktor Dirks.

[www.globaco.de]

Looks to be almost exactly something useful, ie pneumatically operated dispensing syringe with a range of needles and the needles appear to have a locking mechanism that should prevent the needle being blown off if it gets either clogged or the feedstock is a touch on the over viscous side.

Hmmmm options for drive

A larger syringe used as an air pump coupled to a stepper or servo drive, with pressure gauge or switch ?? (Cheap and doable, should be able to use a reprap extruder board to drive this)

Compressed air (Compressor, metering fittings etc expensive if you don't already have it lying around.)

I guess the fact that it is pneumatically operated means that when driven from a stepper the air compression will act as a cushion or spring and should smooth out the perturbations in flow rate caused by stepping the drive syringes plunger.

The actual tool head should be both reprapable or bodged with those funny tool/C clips folk use to hang tools etc from their workshop walls.

Hmmm driving it looks doable, what about stopping the flow an air cushion/spring will continue dispensing until the pressure drops below that that needed to get some of the feedstock to dispense ie great for on poor for quick off. reversing the stepper drive will risk sucking air back into the dispenser (a no no).

Maybe a solenoid vent on the drive side. It would allow a quick off ie release pressure and allow the stepper to reverse and recharge the pneaumatic drive syringe (good if there are leaks) whilst waiting for the next dispense command.


Options & thoughts what do you guys think ????

Viktor any idea how much their items run out at ???? ie cost ??

Cheers

aka47

Hi aka47,

... as they sell in bulks, it's some ten to hundred Euros per order (+shipping), but it's relatively cheap (some ten cents) for single needles and syringes ...

I developed with pressurized-air-dispensers too - it's easy and fast to dispense with PWM or simple pressure-per-speed-calculations, but not so exact in volume and tray-width, as the viscousity/temperature is a very important parameter, which shouldn't vary over the complete fabbing time!

It's a big problem too, if the paste changes consistency over storing time, so you have to adjust the extruding parameters (pressure, PWM-rate, vacuum for stopping and so on) very often.

And low-viscous pastes with bigger needles-diameters tends to rinse and drop down when released and stay a while.

For this a syringe powered by a stepper-motor (as in the fab@home-dispenser) is much more accurate - and there is no problem with moving a short way back to proper stop the paste-flow, when you adapt the parameters for the actual paste (comercial air-powered dispensers do the same with a short vacuum-shot)

Viktor

... as Globaco didn't have an english catalogue, here is another German/European distributor with similar costs, but an english page:
[www.martin-smt.de]

Viktor

Viktor

Are Globco's needles sharp or blunt ie is the end cut at an angle or perpendicular.

Angled or sharp needles are likely to be upset when drawing in the wrong direction. (Like the nib on a writing pen). These are the standard types for medical work as they are generally used to puncture people with.

Blunt or cross cut needles should be fine providing the pen to workpiece distance is correct (Like rotoring drawing pens).

If they are already blunt it could be worth buying in needles from a dispensing outfir like Globco and buying in syringes with a matching locking arangment from a medical suppliers.

Re changes to viscosity.

If a servo direct drive system is used instead of a stepped direct drive system (as opposed to a pneumatic one) it should compensate for a quantity of this. It is also possible to detect when viscosity has gone out of spec if current and/or back current are metered on a servo system.

I must admit direct drive is mechanically simpler.

cheers

aka47

(Going to look for some part time work to tide me over)

... the dispenser-needles are all blunt/perpendicular - you can order them with different lengths or with different bended angles (mostly 45

I was thinking along another path, having done my own pcb's from the age of eleven or so.

If we are still going to use ready made PCB material, why not use ones that are already sprayed with photo resist layer?

Instead of spurting ink we could use a fiber of UV-light going around until the pattern is exposed. resolution should be descent

Saving time would be to design the PCB patterns so that we only need to draw a thin line to isolate areas instead of drawing them.

If I get some time left I will try to make the reprap PCB's using some home-cooked equipment just to see how they turn out.

The whole thing would still have to be etched in FerroChloride or similar, right?

The largest trouble I have had so far is to make the two sides align.

I also saw some fun thread (here?) about some guys using salt water and electrolysis to etch the copper away, thus avoiding the aggressive and unfiendly etch-liquids. The requirement was that all parts of the copper being removed had to be connected in one electric circuit, making the Cad:ing a bitch.

Maybe we could just paint the whole thing, then drawing scratches with a fine pin (no syringe needed) and then drown the whole thing in saltwater and use two electrodes to "fizz" away the copper where needed.

One might even be able to meassure when contact is lost and thus know when the etching is "done" for that part.

Sorry that was about a hundred suggestions in one, comments?

Hi mimarob,

... in past i 'drawed' normal and outline-PCB's with a LED, which i lathed to a sharp tip, covered with black colour and polished the tip with sand-paper, until i had the right sized light-throughput - so i managed to make from old HP-plotter-pens some 'light-pens' with different spot-diameters from ~0,2 to 1 mm.

As then only green LED's were common, i had to search a time, until i found a negative and positive photosensitive spray-lack with sensitivity for green light ...

Later i milled prototypes of outline-PCB's with my CNC-mill and etched small series with chemistry.

AFAIK some PCB-layouting programs already output outline-PCB's - for safe Target ( [www.ibfriedrich.com] , examples: [www.ibfriedrich.com] )
- and maybe the actual Eagle-CAD ( download english version: [www.cadsoft.de] prototyping-tools: [www.cadsoft.de] )...

Both programs have a fully functionable freeware-version (for smaller doublelayer PCB's) and there are not so expensive Non-Profit-licences too ...

Viktor

OK to summarize the freewheeling thinking process so far.

First most importantly thanks to everyone for input.

Assumptions

1. The objectives and methods that follow are aimed at a working machine.
2. To get to a working machine even if only a strap version strip board construction will be the entry level for the initial electronic bits. (you've got to have boots to strap)
3. Having achieved a working machine and PCB production capability PCB designs may then be used to replace the strip board builds and components optionally reused.

Objectives

1. Produce the means to rapidly produce/prototype PCB's using reprap style machinery
2. Do above in a way that minimizes production steps (as much as possible is done in as few a number of operations as possible)
3. Do above in a way that is as eco friendly/safe as possible.
4. Do above in a way that is as easy as possible for an electronic newbie to get to grips with (appropriate technology)
5. Do above in a way that is cost effective. ie inexpensive

Methods - ie potentially viable techniques for PC production

1. Direct draw of etch resist onto bare copper board using one of :-
1a. Drawing pen with some form of ink to be trialled.
1b. Drawing pen made from syringe dispenser with needle nib using real etch resist ink.
1c. Extrude etch resistant plastic tracks
1d. Use cut vinyl cut with a drag knife and transfered. (might not stay stuck during etching)

2. Direct draw of photo mask onto presensitised copper board using one of :-
2a. Drawing pen with some form of ink to be trialled.
2b. Drawing pen made from syringe dispenser with needle nib using high opacity ink.
2c. Extrude photo masking plastic tracks
2d. Use cut vinyl cut with a drag knife and transfered.

3. Conventional UV exposure of photosensitised board following :-
3a Drawing of exposed areas using light pen of appropriate wavelength and collimation.
3b Creation of photo mask film by drawing on photo sensitive film using light pen as in 3a
3C Creation of photo mask film by drawing photo mask on acetate or equivalent using drawing pen with some form of ink to be trialled
3d Creation of photo mask film by drawing photo mask on acetate or equivalent using drawing pen made from syringe dispenser with needle nib using high opacity ink.

4. Mill (route etc) prototypes :-
4a In outline
4b remove all unwanted copper (to mean all not just outline)

Additional nice to have's (may actually be beneficial to objectives to use same/similar tooling as above)

1. Method of drawing ident layers direct on to Boards
2. Method of drawing solder/plating resist layers on to boards
3. Method of dispensing solder paste direct on to boards (for surface mount)
4. Method to drill/mill holes and board shapes


Notes

I have avoided chemistry etc here as it is relatively off topic

Have I missed anything, is there anything else that can be added at this point that is viable ????

Cheers

aka47

Edited 5 time(s). Last edit at 12/06/2007 06:15AM by aka47.

Hi aka47,

... a common method for DYI-PCB-making is to print with laser-toner (=latex and graphite) on film or paper and transfer the printed lines by ironing the film or ironing and water-removing/solving of the paper ...

For best synchronizing two-layered PCB's you can draw 4 fitting-pads and drill them before making the second side, so you can place the PCB exactly with inserting pins in the drilled holes and pass-fitting holes in the mill-bed ...

Viktor

Hullo Viktor

Yup that too..

In have tried this method with very mixed results. Adding more steps into the process creates more things to go wrong with the attendant expense. (Time and Money)

I must admit to have been at this sort of thing for a long time and have tried more methods of PCB Prototyping and production than I care to think about. (Including twice size layout on light box and colored tape). I got into electronics etc when 8 and am now top side of 40.

Overall I find PCB production frustrating (it is a means to an end really) the amount of time effort, black art and chemistry that is expended is massively disproportionate to the end benefit unless you are going into volume production. I would rather be getting on with the Electronics or software.

I guess I can summarize it all as:-

There are 101 ways to skin a cat, some of those are less bloody than others. ><

I think William of Occam had the same ideas in his razor approach

If your in the states I found these folk who do the syringes (manual and air) and dispensing needles.

[www.howardelectronics.com]

Min quantities apply. But individual prices are so ridiculously cheap it almost begs someone to buy some useful sizes and sell them on in hobbyist quantities.

In the UK I have found a number of company's that do the Techon range notably RS and Farnels do them too but don't seem to stock all the sizes.

The question is what are useful sizes ????

aka47

Edited 1 time(s). Last edit at 12/06/2007 06:03AM by aka47.

Something I was reminded of whilst surfing for plotter pens (drawing pen style) as these are designed to handle being dragged across the surface of paper without fine contro of down force.

Vinyl cutter pens for plotters, a friend of mine uses them for doing stick on signs etc for cars and other things.

Here's what Wikipedia had to say about the process.

[en.wikipedia.org]

Here's a couple of hacks for bodging a vinyl cutter from a Hewlet Packard pen plotter (just another cartesian robot really)

[www.luberth.com]

[www.hackaday.com]

I don't know about PCB production (although it looks sort of doable for low res work) but this would be great for Rapid Prototyping of front panel decals & labels etc. Not to mention cutting out reverse printed film front panels etc.

I'll edit the summary post and add the technique to the list.

Cheers

aka47

Personally I think the PCBs (Stepper Motor V1.1 especially) needs to be redesigned to be single layer boards with larger track withs, so that they are more suitable for both home production, and reprap production. It wouldn't hurt to make them larger at this point either. The costs saved from reducign them to single layer board shoudl outweigh the costs of making them larger.

The V1.1 stepper motor pcb looks quite bad when viewed in Eagle, it seems the auto routers done all the track routings and they're not always in the best of places.

I've begun redesigning the board, using the l297/l298 combo, but also using a new stepper motor controlelr from Allegro, which contains their versino of the l297/l298 and the diodes.

That sounds sensible to me..

There is a school of thought that suggests PCB's should have the minimum (permissible for the application, allowing for DRC, inter track Capacitance and Inductance) of copper removed.

1. It means you get more boards out per liter of etching solution and is therefore more cost effective and less polluting.

2. It gives a higher degree of heat sinking on board. more copper = more cm sqr to dissipate heat from.

3. Thicker tracks means less track lift when soldering by hand ergo less repairs and messing around

4. Thicker tracks means less resistive loss and more current carrying capability

5. Making as much unused copper into power ground plane as possible plane reduces the criticality of those 1000pf capacitors you need to sprinkle everywhere to stop logic and switching circuitry loosing it on or around transitions & glitches (sorry forgot the technical name for them) Particularly when double siding a board upper layer + plane with lower Gnd plane can make quite a difference to operational stability. I guess it also reduces EMI levels too.

6. I guess finally if you are etching less copper it all takes less time.

Given that steppers are savagely predisposed to current spikes and resonances of both the mechanical and electrical kind your suggestions sound good.

Hmmm never heard of the newer component you mentioned do you have a part number etc so I could have a nosey just for fun.

cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Umm

I also forgot to mention most importantly that double sided boards are a pain in the rump if not near impossible for the average kitchen table, once a blue flood artiste.

(I am not counting myself out of this category, I have yet to successfully register Double sided prints on my kitchen table)

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

... when in 1986 i got my first CNC-mill, the first try was to go away from etching PCB's - so i hacked a program on my Atari-ST, which could scan for lines in Bitmaps, make contour-outlines and export them as HPGL.

Then i made a driver for the mill, which milled the HPGL-files - this were fine outline-insulated PCB's with big copper-areas between the tracks.

For double-sided PCB's i manually drilled holes in the areas and connected them through and to GND, so the EM-shielding was absolutely perfect

I build my stepper-driver-boards and the interfaces and powersupplies to a RF-CO2-Laser with this method and could eliminate all spikes and EM-sensitivities (i'll post some images, when at home).

---

In a further experiment i tried to ablate the copper with a NdYAG-laser, what was not so easy, as copper is very slow to ablate and then the PCB 'explode' when you cut trough the last copper.

I solved this with 75% thicker copper on the PCB's with only ablating 80% of the copper and then put it in the etch-liquid to etch down 30%-40%, so the pre-engraved trays were completely etched and all other areas were etched too and thinner then at beginning, but thicker as normal PCB's ...

Viktor

I think that there is a quantity of useful information collected in this thread.

Would it possibly be better placed being moved into somewhere suitable in the Wiki ??

I.E. quicker and easier to find with a link back to this thread so folk can follow the thinking if they want.

I am thinking of the summary and possibly adding a comparative table taking into account the strengths etc of the current Darwin model to help folk decide what they might want to have a go at now versus look at in preparation for the "Next Generation".

What is the form with the Wiki for this group, is it "by invitation" or "register and wait" or something else ?????

I currently use Twiki on my home website and have got enough time at the moment to make a contribution here.

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

The chip is an Allegro SLA7052M. Have a look here: [www.allegromicro.com]

Essentially most of the components are internal, it's a uni-polar driver, and it seems most people don't have these. I can't remember whether it can be converted easily. It just needs a Vref, sense resistors, smoothing cap and inputs.

I am using different motors than the "official" stepper motors sold/used by most reprappers. I got them for free tho, mm salvage!

Whats wrong with using an inkjet printer head directly on the rep-rap? Or ripping up a whole printer?

[techref.massmind.org]

Seems the optimal method - I dont know what is involved with making an inkjet printer head work without the rest of the printer though

J.