Why isn't UV Laser based PCB manufacturing more mainstream yet?

Thread Starter

vidit_1248

Joined Dec 17, 2020
7
While feeling the shared pain of costly and slow PCB Manufacturing process. Which also makes the hardware creativity and education much costly and slow.
I was trying to figure out how can we decrease the cost of PCB manufacturing, get it made quickly and still maintain the level of quality we currently get. I came across the concept of using UV laser for Board cutting boards and drilling board plus using Masked SLA type of process for UV expose.

To me, this seems like a revolutionary setup.
- You can make every board different, which makes enables mass customization
- Very less wear outs as there are no drill bits to die
- Automation can be a much higher decreasing need for costly labour and faster lead time
- Fewer machines and setup costs which makes the footprint of the plant smaller
- Above 2 points can make PCB Manufacturing feasible outside china as well
- Maybe we can mitigate the setup cost of PCB and bring down the cost of proto PCBs to the cost of PCB manufactured at scale

All these seem that for at least 2-layer rigid or flexible PCB, a lot of initial costs can be mitigated and prototype PCB manufacturing cost can be decreased.

So what am I missing? What are the technical or commercial challenges you see in making with this way of manufacturing a PCB?
 

jpanhalt

Joined Jan 18, 2008
11,087
Welcome to AAC.

Have you tried subtractive laser for making a PCB? How did copper ablation work with fine detail? Drilling holes? What about the burns? Some hobbyists have tried laser for ablating an etching mask or exposing a photoresist etching mask with mixed results.

The biggest problem I see is scaling from single boards to whole panels to lots of panels. With your laser idea, every board has to be drawn, regardless of how many. Thus, the time for a full panel is dependent on the amount of circuitry on that panel. With printing, the exposure time for a full panel doesn't increase the same way.

Finally, modern PCB manufacture uses additive processes ("pattern plating"), not just subtractive. Search for videos of how boards are currently made. Here's one description: https://www.allflexinc.com/blog/plating-process-options-for-flexible-circuits/ How will your laser idea decrease those costs? Just guessing doesn't count.
 

Thread Starter

vidit_1248

Joined Dec 17, 2020
7
Thanks for the welcome and reply!

Ya, I agree ablating copper using a laser is not a good idea.
But how about using transparent LCD in place of transparent film and then etching the conventional chemical method?
This keeps scalability intact. while taking away film printing step (plus cost) and make the design easily changeable at the same time.
Many people seem to have tried it by hacking into their DLP printers and have got pretty good results, so I think maybe engineering that process well specifically for this purpose might give results which are equal quality or better quality compared to an existing process, what do you think?

Thanks for pointing at pattern plating, that's something I didn't know.
 

jpanhalt

Joined Jan 18, 2008
11,087
Thanks for the welcome and reply!

Ya, I agree ablating copper using a laser is not a good idea.
But how about using transparent LCD in place of transparent film and then etching the conventional chemical method?
This keeps scalability intact. while taking away film printing step (plus cost) and make the design easily changeable at the same time.
Many people seem to have tried it by hacking into their DLP printers and have got pretty good results, so I think maybe engineering that process well specifically for this purpose might give results which are equal quality or better quality compared to an existing process, what do you think?

Thanks for pointing at pattern plating, that's something I didn't know.
Work out the details of your ideas. Is your LCD idea something like the current, inexpensive, 3D printers that use an LCD mask? If so, you need to analyze the steps in that and compare to the current process. Will you get close to the same resolution?
 

Sensacell

Joined Jun 19, 2012
3,432
Just thinking out loud here.

IMHO- If you invest the time and energy to do a PCB layout using software, the value of making your own board isn't there.
You will have no vias, no plate-thru holes, and you still need to drill all those holes.
The end result is not good- and you need to create your own Love Canal chemical disaster area in your own home.

Milling them is even worse! listening to the spindle scream for an hour, to find that the tool snapped, or the board was not flat so it only cut half the board.

I think the sweet spot would be a mechanical method that cut the copper (no chemicals) that was not milling (too slow and noisy) that could make quick and dirty boards.
Of course you would have to design for the process, it's not the same as a traditional board, but it could allow faster prototyping.

What I imagine is some kind of "CNC swivel knife" kind of machine, with a carbide cutter that would rotate as it moved, cutting groves into the board.
Much faster and quieter than a spinning cutter.
 

jpanhalt

Joined Jan 18, 2008
11,087
What I imagine is some kind of "CNC swivel knife" kind of machine, with a carbide cutter that would rotate as it moved, cutting groves into the board.
Much faster and quieter than a spinning cutter.
Can we anticipate a prototype here in 2021?

I don't think you have considered the matters of speed, noise, flatness, variation in copper thickness, and removed scrap thoroughly. Do you have experience with the type of rotating knife you describe? Or put another way, a V-shaped mill is a rotating knife.
 

Sensacell

Joined Jun 19, 2012
3,432
Can we anticipate a prototype here in 2021?

I don't think you have considered the matters of speed, noise, flatness, variation in copper thickness, and removed scrap thoroughly. Do you have experience with the type of rotating knife you describe? Or put another way, a V-shaped mill is a rotating knife.
The knife does not "spin" - it rotates to the vector of movement.

What I am imagining is "automating" the process shown in the video below.

 

Thread Starter

vidit_1248

Joined Dec 17, 2020
7
Work out the details of your ideas. Is your LCD idea something like the current, inexpensive, 3D printers that use an LCD mask? If so, you need to analyze the steps in that and compare to the current process. Will you get close to the same resolution?
Considering people are been able to get 4 mills just by hacking 3D printer (That too DLP), I think if we engineer it well using a 13 inch LCD of 4K resolution we can get masking of 3 mills (0.075 mm).
How I Imagine the pattern applying process:
1. Apply the UV sensitive film on the Board using an automated process
2. Place the board with UV Sensitive film below the LCD
3. Bring the PCB pattern on LCD
4. Turn on UV panel on the other side of LCD
5. Wash away unwanted UV Sensitive film
 

Thread Starter

vidit_1248

Joined Dec 17, 2020
7
Just thinking out loud here.

IMHO- If you invest the time and energy to do a PCB layout using software, the value of making your own board isn't there.
You will have no vias, no plate-thru holes, and you still need to drill all those holes.
The end result is not good- and you need to create your own Love Canal chemical disaster area in your own home.

Milling them is even worse! listening to the spindle scream for an hour, to find that the tool snapped, or the board was not flat so it only cut half the board.

I think the sweet spot would be a mechanical method that cut the copper (no chemicals) that was not milling (too slow and noisy) that could make quick and dirty boards.
Of course you would have to design for the process, it's not the same as a traditional board, but it could allow faster prototyping.

What I imagine is some kind of "CNC swivel knife" kind of machine, with a carbide cutter that would rotate as it moved, cutting groves into the board.
Much faster and quieter than a spinning cutter.
Ya I am also saying let's get rid of Drill Bits and saws etc, just use lasers to do drilling and cutting, no noises at all!
And of course no drills to break and die on us.
 

jpanhalt

Joined Jan 18, 2008
11,087
Considering people are been able to get 4 mills just by hacking 3D printer (That too DLP), I think if we engineer it well using a 13 inch LCD of 4K resolution we can get masking of 3 mills (0.075 mm).
How I Imagine the pattern applying process:
1. Apply the UV sensitive film on the Board using an automated process
2. Place the board with UV Sensitive film below the LCD
3. Bring the PCB pattern on LCD
4. Turn on UV panel on the other side of LCD
5. Wash away unwanted UV Sensitive film
Now you need to try it. Simple.
 
I am not expert enough to determine whether your idea has merit or not.
But IF it does, I would not be discussing it on an open forum.
Unless you want to find out later that a large corporation has started selling machines based on your ideas.
 

shortbus

Joined Sep 30, 2009
10,045
I worked where they made wire, terminals and wire harnesses. When prototyping a new terminal they would make a strip of blanks made from brass or copper in the shape of the terminal blank to make sure all of the sizes were correct before making the dies. They had a Trumph laser cutter, and when they tried cutting the brass or copper it didn't end up working very well. Seems that both materials conduct the heat away and reflect the laser light. They ended up wire EDMing the shapes.
 

Thread Starter

vidit_1248

Joined Dec 17, 2020
7
I am not expert enough to determine whether your idea has merit or not.
But IF it does, I would not be discussing it on an open forum.
Unless you want to find out later that a large corporation has started selling machines based on your ideas.
This Idea has already been in an open forum (not totally my idea, to be frank).
And still, there is no solution out there, hence I am wondering the issues with it
 

Thread Starter

vidit_1248

Joined Dec 17, 2020
7
I worked where they made wire, terminals and wire harnesses. When prototyping a new terminal they would make a strip of blanks made from brass or copper in the shape of the terminal blank to make sure all of the sizes were correct before making the dies. They had a Trumph laser cutter, and when they tried cutting the brass or copper it didn't end up working very well. Seems that both materials conduct the heat away and reflect the laser light. They ended up wire EDMing the shapes.
Oh okay, thanks for that insight. That might be one of the reasons. Do you remember what kind of laser was it? (Diode, CO2, etc)
 

Deleted member 115935

Joined Dec 31, 1969
0
Simple answer,
may be to simple,
cost.

Existing setup works for most people,
existing setup is cheap
existing setup has already purchased the equipment so they want to maximise its profits
qed. Any new technology would have to be an order of magnitude cheaper to break in on cost,
and new tech is not cheap.

Would love one myself,
but it aint going to become cost effective unless shipping becomes very slow and expencive.
 

Thread Starter

vidit_1248

Joined Dec 17, 2020
7
Simple answer,
may be to simple,
cost.

Existing setup works for most people,
existing setup is cheap
existing setup has already purchased the equipment so they want to maximise its profits
qed. Any new technology would have to be an order of magnitude cheaper to break in on cost,
and new tech is not cheap.

Would love one myself,
but it aint going to become cost effective unless shipping becomes very slow and expencive.
Of course, that's always a barrier for new tech.
I guess it has to compete in things other than cost that way
 
Top