micro scraper geometry and scraping etch resist geometry

[coppercone2]

I was wondering about very precise etching, and one problem is the etch mask, but the other problem is the definition of the photoresist layer.

For simple PCB, we currently can use high speed end mills to machine copper.

But I thought of a combination process, that is using etching, but instead of photoresist, or transfer, or being reliant on printers or projectors, to use a scraper needle.

For something simple like a planar coil or other planar RF/MW structure, I was wondering what kind of 'exposure' you can get from using a needle type scribe and very strait rails as a reference to drag the needle on.


The work would be conducted under the microscope, and you might need some kind of compliant system for the scraper so it does not dig into the substrate too much, like controlled pressure by some means.

Scratches can be very fine.

I am not sure about the tool geometry, or materials choice, perhaps a single diamond crystal.


Also the mask would have to be probobly tailored to scraping. Low self healing effects.


I thought someone might find this interesting.


Kind of something like a defraction grating manufacture ?. I thought working at low temperatures might ensure proper 'removal' of the coating.

[Doctorandus_P]

You can attempt to scrape just though some solder resist layer, but you have to be careful not to damage the copper (too much).
Or you go the other way, and just scrape thought the whole copper layer.



It looks like you could get quite good results with that method if you do a few passes with a small offset.

[coppercone2]

it makes me wonder if you can build a little car that drives over the PCB scratching it, like 4 wheel independent drive. it should work if its heavy enough


hmmm it seems like a really powerful process especially if you fit it with piezo positioner like a hard drive head... put a piezo positioner on the end of a milling machine with a sharp scribe...

that should be capable of drawing incredibly fine geometry ?


also, who cares about damaging the copper layer? etchant will eat that, so long it does not disturb the nearby coating. I am afraid it will tear and drag and stuff like that.


and im really damn curious if ultrasonic might help here. I Have that probe. I think I will make a collet to attach a carbide scribe to the end of the ultrasonic wand and see how it does, thats actually a achievebale project. maybe it would prevent the stupid little things happening that I am predicting

[jpanhalt]

Why not go all the way and use EDM?

[coppercone2]

I don't know about that, I followed a few people on youtube trying EDM and it looked like a total god damn nightmare.



Though, it usually seemed like it was some problem with the brass roller shit (band saw type). I don't know anything about electrode EDM. Maybe that one is achievable. I know that the band saw type EDM is not for me.

[jpanhalt]

I don't think YT is a very good way to evaluate anything.  EDM is used industrially every day for precision machining.

[coppercone2]

there is some serious engineering that can't easily be done by outsiders in that process

[coppercone2]

what if you added a plow to this bad boy?
https://www.nbcnews.com/id/wbna9778004

maybe it can make a few atom wide groove? get rid of all the ultra ultra violet crap they are using now to make transistor masks

Maybe it can drag a single atom of plutonium or something that weakens some kind of resist

[coppercone2]

for instance, EDM requires the use of a pressure washer, the simple drill EDM as shown by that guys science channel.

Something about having to hook in a pressure washer into a machine tool just makes me not wanna do it. totally ridiculous.

unless there is some more tame EDM technology, perhaps there is, for such light work.