Homebrew PCBs

Hi all,

Has anyone here looked seriously at the use of conductive and insulating
inks in an inkjet printer?

I'm doing PCBs now which are more and more needing 2 or more layers, and
facing the choice of building up PTH equipment or soldering wires
through vias, neither of which are particularly appealing.

It occurred to me that it might be worth looking into the possibility of
a colour inkjet printer with a conductive ink in one of its CMY tanks,
and an effective insulating ink in another tank.

Then, a PCB of n layers could be printed onto a thin substrate (eg
inkjet gloss paper) which could then be glued to blank fibreglass, as
follows:
 - one pass with conductive ink, to print all the pads as well as a
   first layer of traces
 - one pass with insulating ink, to cover the areas of the
   previous conductive printing that will be covered by the next
   conductive layer
 - second layer of conductive ink
 - second pass of insulating ink, to prepare for the 3rd layer of
   conductive ink
 - ...

The challenges I see with this are:

 - finding an affordable, highly conductive ink with particles fine
   enough to pass through the inkjet nozzles, and print without errors,
   and without clogging the nozzles

 - finding a good insulating ink that will get enough coverage of enough
   thickness to avoid shorts or other electrically undesirable effects

 - converting n-layer PCB artwork files into a set of alternating layers
   of conductive and insulating ink

 - ensuring accurate registration of the substrate through n passes 
   through the printer - rotational error would need to be pretty much
   zilch, x and y errors would need to be around 4mil or less

Anyone got any thoughts on this?

Cheers
David

I've been wondering about this.

There's a compound of Gallium, Indium and Tin that freezes at minus 4  
degrees F.    So it's liquid at room temperature.

You can coat your own mirrors, paint it on stuff, maybe you could use  
it to draw antenna grids on glass.

If you use pure Gallium, it would melt in your hand, but freeze below  
that temperature to a solid.    Even if you could get pure Gallium to  
work in an inkjet you'd need a heater to keep it warm or it would  
freeze up in the ink tank.

William Carr

On Feb 9, 2007, at 3:18 PM, David McNab wrote:

> - finding an affordable, highly conductive ink with particles fine
> enough to pass through the inkjet nozzles, and print without errors,
> and without clogging the nozzles
>
> - finding a good insulating ink that will get enough coverage of  
> enough
> thickness to avoid shorts or other electrically undesirable effects
>
> - converting n-layer PCB artwork files into a set of alternating  
> layers
> of conductive and insulating ink
>
> - ensuring accurate registration of the substrate through n passes
> through the printer - rotational error would need to be pretty much
> zilch, x and y errors would need to be around 4mil or less
>
> Anyone got any thoughts on this?
>
> Cheers
> David



[Non-text portions of this message have been removed]

> 
> > - finding an affordable, highly conductive ink with particles fine 
> > enough to pass through the inkjet nozzles, and print 
> without errors, 
> > and without clogging the nozzles
> >
> > - finding a good insulating ink that will get enough coverage of 
> > enough thickness to avoid shorts or other electrically undesirable 
> >
> > Anyone got any thoughts on this?
> >
> > Cheers
> > David
> 
> There's a compound of Gallium, Indium and Tin that freezes at 
> minus 4  
> degrees F.    So it's liquid at room temperature.


Or find an etchant that works at low temperatures (and only attacks the
copper, of course).  It would make a change to toss something in the fridge
rather than into a tub of warm water.

This would be something a Peltier module can help with.  It can keep the PCB
cool, and be cheap and not take up much space while doing it.

Tony

On Sat, 10 Feb 2007 07:20:03 +0100, Tony Smith <ajsmith@...>  
wrote:

>
> Or find an etchant that works at low temperatures (and only attacks the
> copper, of course).  It would make a change to toss something in the  
> fridge
> rather than into a tub of warm water.
> This would be something a Peltier module can help with.  It can keep the  
> PCB
> cool, and be cheap and not take up much space while doing it.
> Tony


Only that usually chemical reactions speed up with increasing temperature.
Therefore i doubt such an etchant exists.



ST

On Sat, 10 Feb 2007 06:44:25 +0100, William Carr <Jkirk3279@...>  
wrote:

> I've been wondering about this.
> There's a compound of Gallium, Indium and Tin that freezes at minus 4
> degrees F.    So it's liquid at room temperature.
> You can coat your own mirrors, paint it on stuff, maybe you could use
> it to draw antenna grids on glass.
> If you use pure Gallium, it would melt in your hand, but freeze below
> that temperature to a solid.    Even if you could get pure Gallium to
> work in an inkjet you'd need a heater to keep it warm or it would
> freeze up in the ink tank.
> William Carr


I would expect it will have a high surface tension, seeing that it is a  
metal and all. So it may be tricky to keep it in place.
Also, a PCB that melts away if it gets hot is kinda useless for me.

ST

> > Or find an enchant that works at low temperatures (and only 
> attacks the
> > copper, of course).  It would make a change to toss 
> something in the  
> > fridge
> > rather than into a tub of warm water.
> > This would be something a Peltier module can help with.  It 
> can keep the  
> > PCB
> > cool, and be cheap and not take up much space while doing it.
> > Tony
> 
> 
> Only that usually chemical reactions speed up with increasing 
> temperature.
> Therefore i doubt such an etchant exists.


Probably.  I doubt FeCl would work in the fridge, or it'd take a week to
work.

Anyway, 'high' temperature alloys exist, like Woods metal.  It melts at
around 50C, so that would work.  Only a couple of trivial problems remain,
like is Woods metal attacked by etchant, and how does one pump molten metal
onto a PCB?

Hmm, PCB sits on a Peltier, cooled enough to avoid condensation.  Molten
metal held in a syringe, forced out by plunger, retract plunger slightly to
stop.  Same as solder paste.  If you test conductivity between the PCB & the
shrine, you can see when the track start to get laid.  How to do wider
tracks?

PCB is etched, then placed in hot water to remove Woods metal.  Metal is
reused.

Right then, "who wants to try it", he says, thus showing an idea is worth
zero unless you can show it works.

Tony

On Sun, 11 Feb 2007 12:47:34 +0100, Tony Smith <ajsmith@...>  
wrote:

>
> Right then, "who wants to try it", he says, thus showing an idea is worth
> zero unless you can show it works.
> Tony


And nobody volunteers to try an idea that is so unlikely, and would if  
everything works out just great, still be more complicated and expensive  
than most existing processes.

ST

> > Right then, "who wants to try it", he says, thus showing an 
> idea is worth
> > zero unless you can show it works.
> > Tony
> 
> 
> And nobody volunteers to try an idea that is so unlikely, and 
> would if  
> everything works out just great, still be more complicated 
> and expensive  
> than most existing processes.
> 
> ST


Oh, c'mon, where's your sense of adventure.  Can't be any more complicated
than milling PCBs, building a spray etcher, a CNC router for drilling,
solder paste dispenser or most of the stuff that happens on these lists.

If anyone out there has some Woods metal, stick it on a PCB, etch it and
tell us what happens.

Hell, it might even work.

Tony

Woods metal is an alloy of tin-lead-bismuth-cadmium. Hmm, low 
temperature solder. I suppose that is one way of tinning your boards :-)

Les

> Anyway, 'high' temperature alloys exist, like Woods metal.  It melts at
> around 50C, so that would work.  Only a couple of trivial problems remain,
> like is Woods metal attacked by etchant, and how does one pump molten metal
> onto a PCB?
>
> Hmm, PCB sits on a Peltier, cooled enough to avoid condensation.  Molten
> metal held in a syringe, forced out by plunger, retract plunger slightly to
> stop.  Same as solder paste.  If you test conductivity between the PCB & the
> shrine, you can see when the track start to get laid.  How to do wider
> tracks?
>
> PCB is etched, then placed in hot water to remove Woods metal.  Metal is
> reused.
>
> Right then, "who wants to try it", he says, thus showing an idea is worth
> zero unless you can show it works.
>
> Tony
>

> Woods metal is an alloy of tin-lead-bismuth-cadmium. Hmm, low 
> temperature solder. I suppose that is one way of tinning your 
> boards :-)
> 
> Les


Unless you pass too many amps thru it!

A quick rummage around the 'net turned up Wood's metal is 50% bismuth, 26.7%
lead, 13.3% tin, 10% cadmium by weight.  I don't think that's RoHS
compatible.  It melts at 70C, not 50C as I thought.  Someone once told me
that when pointing out it was used in fire sprinkers.  Might have been a
different alloy.

However, there's Field's metal, 60C melting point and none of the nasty
stuff, being 32.5% bismuth, 51% indium, 16.5% tin by weight.

There's a bloke on eBay selling a 58% Bismuth, 42% tin alloy, melts at 290F.
That's a bit high.

Years ago I turned down the chance to buy a large chunk of Field's metal,
dang.  It comes in handy for casting and weird machining tricks.  To machine
something thin, funny shaped, delicate or otherwise annoying, you cast it
into the Field's metal, machine it, then melt it off.  Since it's supported
by the other metal, the piece you're machining doesn't get damaged.

Tony

On Sunday 11 February 2007 09:19, Tony Smith wrote:
> > Woods metal is an alloy of tin-lead-bismuth-cadmium. Hmm, low
> > temperature solder. I suppose that is one way of tinning your
> > boards :-)
> >
> > Les
>
> Unless you pass too many amps thru it!
>
> A quick rummage around the 'net turned up Wood's metal is 50% bismuth,
> 26.7% lead, 13.3% tin, 10% cadmium by weight.  I don't think that's RoHS
> compatible.  It melts at 70C, not 50C as I thought.  Someone once told me
> that when pointing out it was used in fire sprinkers.  Might have been a
> different alloy.
>
> However, there's Field's metal, 60C melting point and none of the nasty
> stuff, being 32.5% bismuth, 51% indium, 16.5% tin by weight.
>
> There's a bloke on eBay selling a 58% Bismuth, 42% tin alloy, melts at
> 290F. That's a bit high.
>
> Years ago I turned down the chance to buy a large chunk of Field's metal,
> dang.  It comes in handy for casting and weird machining tricks.  To
> machine something thin, funny shaped, delicate or otherwise annoying, you
> cast it into the Field's metal, machine it, then melt it off.  Since it's
> supported by the other metal, the piece you're machining doesn't get
> damaged.
>
> Tony

Tony,
Smallparts.com sells low temperature alloys of that kind.
http://www.smallparts.com/products/descriptions/lma.cfm

My group has been experimenting with printing circuit board traces on plastic, 
if people are really interested.
http://blog.reprap.org/
http://www.reprap.org

Regards,
-Sebastien Bailard
RepRap.org - self-reproducing 3D printer project.

>> > Woods metal is an alloy of tin-lead-bismuth-cadmium. Hmm, low
>> > temperature solder. I suppose that is one way of tinning your
>> > boards :-)
>> >
>> > Les
>>
>> Unless you pass too many amps thru it!
>>
>> A quick rummage around the 'net turned up Wood's metal is 50% bismuth,
>> 26.7% lead, 13.3% tin, 10% cadmium by weight.  I don't think that's RoHS
>> compatible.  It melts at 70C, not 50C as I thought.  Someone once told
>> me
>> that when pointing out it was used in fire sprinkers.  Might have been a
>> different alloy.
>>
>> However, there's Field's metal, 60C melting point and none of the nasty
>> stuff, being 32.5% bismuth, 51% indium, 16.5% tin by weight.
>
>
> Smallparts.com sells low temperature alloys of that kind.
> http://www.smallparts.com/products/descriptions/lma.cfm
>
> My group has been experimenting with printing circuit board traces on
> plastic,
> if people are really interested.
> http://blog.reprap.org/
> http://www.reprap.org
>
> Regards,
> -Sebastien Bailard
> RepRap.org - self-reproducing 3D printer project.


Looking at some of those prices make me wonder what temperature gold melts
at - it might be cheaper.  :)

Tony