PCB photoresist printer

Discussion in 'General Electronics Chat' started by praondevou, Jan 13, 2016.

  1. praondevou

    Thread Starter AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Has anyone successfully be able to print high density/small package/fine pitch PCB footprints with a converted laser printer? There are several people who seem to have been able to print directly on a PCB with converted printers but this doesn't seem to be really good for small packages. I'm talking about packages as small as SOT-953 or fine pitch ICs...

    Thanks
     
  2. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,026
    3,789
    I haven't seen anyone successfully print directly to copper with a laser printer (inkjet yes, laser no). I thought the laser needed to create a charged surface on the toner drum and that charge had to remain as it comes in contact with the printed substrate (paper or film) as it is melted to the surface. As an excellent conductor of heat and electricity, copper dissipates the melting temps and charge quickly and toner does not adhere and gets the image on the toner drum gets blurred because of charge dissipation. If you have evidence to the contrary or a trick to overcome this issue, I would love to see how it is done.

    On top of that, you either need flexible PCB films or a straight-pass printer - only a few models have a straight-though feed system for rigid stock.
     
    Last edited: Jan 13, 2016
  3. praondevou

    Thread Starter AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    There seem to be people that did it with straight-through laser printers, don't know how they managed to keep the copper positively charged...
    I'm more worried about the superfine pitch I will be trying to do.

    Inkjet could be ok too, there are some good examples here: http://techref.massmind.org/techref/pcb/etch/directinkjetresist.htm
     
  4. RichardO

    Well-Known Member

    May 4, 2013
    1,235
    384
    Not really n answer to your question but...

    One alternative is to start with a solid layer of resist on the copper and selectively remove the resist with a Laser. You would then have to do a traditional etch of the exposed copper.
     
  5. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,685
    900
    Your premise is very confusing. At first, you question about a photo-resist. For DIY users, that is usually not printed directly onto the pre-sensitized board. Typical photo-resists are insulators, and there should be no problem printing directly to them with a "laser printer," assuming your laser printer has a straight path and the resist can withstand the fusing temperature. (Which I doubt.) But then, you give very few specifics. Words, such as "small" are meaningless. "Small" to a roofer is something something considerably more than an inch. To a machinist, it might be three orders of magnitude smaller.

    I have used photo-resists to reliably make features and separations of about 0.010 inch. I have not tried any smaller, but suspect going to 0.006" would be possible.. The one dimension you do give (i.e., SOT-953) requires ) 6 mil resolution. I think that is possible, but have not made any DIY boards that required it.

    John
     
  6. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,026
    3,789
    below 10 mils (0.010") is quite a challenge with photo-resist but I have done it both with traces and gaps using overhead transparencies and a laser printer. Anything smaller than 10 mils is kind of lucky - especially if your 10 mil traces are long runs. I have done 6 mil traces but I've also had 8 to 10 mil traces fall apart.

    Thin traces requires very high contract (photo-stat) image for transfer with a low-UV absorbing glass to hold the image flat to the PCB to prevent undercutting the photo-resist, rapid etching to limit chemical undercutting (heated tank, agitation/bubbles) and slight dilution of ferric chloride to reduce viscosity and improve mass transfer.

    The problem with all of that is, the photo-stat costs $20 - $30 each and, with board and chemicals, I may as well send out for a board. I've decided 10 mils is my limit unless I need to sneak a short trace between some pins here or there.

    I gave up on toner transfer after my first attempt at photo-resist. Photo-resist is sooo easy and reliable if you have room to leave things set up. Good thing I don't use the bar in my basement - that is my photo-resist PCB area. I use the tub in my garage for etching.
     
  7. RichardO

    Well-Known Member

    May 4, 2013
    1,235
    384
    I once worked with a fellow that had a machine that did custom PCB's. He fed copper clad flex circuit material into the machine. The machine printed the circuit pattern on the copper and then did the etch. He then took the etched flex circuit and glued it to a 1/16" thick fiberglass panel using a sheet of thin double face tape. For 2-sided boards he had to glue a second flex circuit onto the other side of the panel.

    The panel with the flex circuit(s) on it was then placed into a CNC drill to add the holes. I don't remember how he maintained precise alignment through the entire process.

    At this point he had special 1/16" long "tubes" that had a rough texture on the outside. For plated through holes he had to press a tube into every hole that needed a through-the-board connection for a component lead. For IC's he used sockets with gold plated wire-wrap pins to make the connection between the top and bottom layers so he did not have to use the tubes for IC pins.

    Quit a complicated process but he was able to deliver prototype assemblies to me for testing. He cautioned me not to do any soldering on the boards and that I should have him do any changes to the circuit that required soldering. I needed a change and did not want to wait so I did the soldering. Even though I was very skilled at soldering, I still damaged the traces from heat. :( I obviously should have taken his advice.

    This has been pretty long winded but I hope it helps someone learn what I learned but in a less painful way. :)
     
  8. RichardO

    Well-Known Member

    May 4, 2013
    1,235
    384
    A friend was bragging that he could do traces as small as 4 or 5 mils, if he worked at it, using toner transfer. :eek:

    My reply was: "Well that's nothing, I can easily do 1 and even 0 mil traces". :D
     
    GopherT likes this.
  9. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,685
    900
    There is no question that "photoresists" in theory can do very fine detail limited by the wavelength of light used. The question is whether an amateur using a typical laser printer (2000 dpi = 0.5 mil = 12.7 micro mm) can do better than a few "mil." When you get to that level of precision, there are significant differences between FeCl3 and hydrogen-peroxide-based etchants, as well as the effects of the thickness of the copper. None of those variables were defined by the TS.

    I have a feeling this thread is more about daydreaming than reality.

    John
     
    GopherT likes this.
Loading...