Investigations regarding DIY photomasks

[Emperor-TK]

I thought it might be nice to do a follow-up to some assertions I made on a previous thread regarding DIY phototools:

http://www.groupdiy.com/index.php?topic=4373

Sorry if this post breaks any of the image size rules, etc. I'm still HTML challenged. Please let me know if there's anything that needs fixing.

What I did was to examine printed photomasks with each of the standard office printers at my disposal and compare them for opacity. My previous experience has been that no office printer comes close to 100% opacity. Secondly, I was curious to see if my guess that there would be more ink/toner at the line edges was correct. To do the test, I used standard overhead transparency films, each designed for the particular printers (laser or ink-jet). I then photographed the prints at 400x magnification using a laboratory grade (Meiji) microscope with backlighting. Polarization filters were used for contrast enhancement. The lighting settings were held constant for each photograph. The photograph lines were nominally 1 millimeter wide. A couple of the results surprised me.

The first image shows an HP Deskjet 950c inkjet on standard quality. Rather than depositing more pigment at the edges, less pigment was deposited as a result of dewetting of the substrate:

The next image shows the same printer on High quality mode. It appears that the printer used CMY inks to double up on the print. There also appears to be some bleed at the edges due to extra ink deposited:

This image is from an HP2100 1200DPI laser printer set to highest quality. This kind of porosity is typical in my experience. Note that the image is actually smaller than the nominal 1mm line width in the artwork:

HP3100 laser, 600DPI. Note the finer edge definition and smoother coverage with the lasers, but the poorer fill/opacity.

This print surprised me a bit. It is an HP 4550 color laser, which is what I always use to make my photomasks. A new toner cartridge has recently been installed, and it is spewing out toner like it has never before. I normally would see better coverage than the other laser printers, but never the near 100% that is shown here. However, as a side effect to the better coverage, the printed lines are spreading out significantly from the artwork. A grid of 200 micron lines/spaces that I printed became 350 micron lines with 50 micron spaces:

Finally, as a comparison, this is a photomask professionally printed onto PET film by Sefar America/MEC (different artwork). Not much comparison:

Regarding the buildup of optical density at the photomask edges, I stand corrected. I am also now less comfortable with my statement that 100% density is unlikely with a stander laser printer. Still, I think these photos show the potential problems with overexposing boards. Hopefully this helps all the DIY etchers out there.

Regards,
Chris

 

[Svart]

that's great work!

I use a 4550 HP too and it works pretty good. i've also been using a lexmark laser printer(model unknown). i've noticed that if i use the lexmark in a lesser quality mode i get a smoother, but not as thick spread of toner. I just use less exposure and it's all good.

 

[ebeam]

I've been doing the transparency thing for a similar photolithography process. The experience around here is that for larger features (>500 micron or so) most printers do OK and to combat the continuity problem most people just stack several transparencies. For smaller features, even supposed 1200dpi printers can barely pull off true 300 dpi (~75 micron features). The best I found was an older HP 6MP laser printer, but even this one didn't work very well. There is a local print shop that will print 9x12" transparencies for $9 (or 2 for $15). Supposedly they can get down to 8 micron, but the 75 micron features I had printed look excellent. They are called the Type Studio ([email protected]).

 

[PRR]

> To do the test, I used standard overhead transparency films

I'm going to guess that inkjet on paper will be different from inkjet on plastic. As you say, the plastic de-wets, and paper wicks ink, so the ink may end up undersized on plastic and oversized on paper. Also the porosity should be less on paper.

I'm wondering what the effect is on xerography (laserjet). Charge could leak into the white area, or drain into the black area. The errors you show are around +/-10% on a 0.040" line, comparable to a trace passing between pads of a 0.100"-pitch DIP, and +/-10% should not be a problem.

The porosity is interesting. In the intended use, eyeballed paper, holes are no problem as long as they are smaller than the eye can see and the sum of the holes does not dilute the black to grey. In etching, a "grey" coverage reduces conductor size, and there are some largish areas of white that could etch to very poor conductors.

Thanks for the micrographs.

> image size rules, etc. I'm still HTML challenged.

These images are fine. Actually they are just 320 pixels wide, and Ethan's edict is 400-wide max. You could post them a little larger. I'm not sure anything would be gained in this case. (Actually you could crop-off the corners and maybe double the size of the fine detail. But I think the detail presented is enough to know if "this mask will etch nice" or not.)

Ethan gimmicked the forum software so over-size pictures are rendered 400-wide max, with a click to see the image in full oversized glory. So (in theory) you can't put a too-wide picture in-line. The gimmick is that readers' browsers still fetch the full size image file which can be slow for huge pix on dial-up.

 

[gyraf]

Thanks, Chris - good work, useable info!

:thumb:

Jakob E.

 

[Emperor-TK]

[quote author="PRR"]

I'm going to guess that inkjet on paper will be different from inkjet on plastic. As you say, the plastic de-wets, and paper wicks ink, so the ink may end up undersized on plastic and oversized on paper. Also the porosity should be less on paper.

[/quote]

This is fun, I rarely get to talk about this stuff outside of work (I formulate inks for printable electronics). I'm would definitely expect different behavior on most papers vs. film. The issue with paper is that there are six bazillion different kinds of paper out there, with different surface chemistries, porosities, and surface roughnesses, etc. Whenever a potential customer asks "can I print your inks on my paper", my response is "send me your paper, I'll make an ink that works with it". Some papers behave identical to films, some behave identical to sponges. For a real high quality gloss paper, an ink would never even see a wood fiber, but rather a calcium carbonate/polymer composite on the surface. Normally, it would be very tough to print an aqueous ink onto a polymer surface. Surface roughness is actually part of what allows inkjets to print on film. The roughness raises the apparent surface energy of the film, alowing the water-based ink to propperly wet it.

I'm wondering what the effect is on xerography (laserjet). Charge could leak into the white area, or drain into the black area. The errors you show are around +/-10% on a 0.040" line, comparable to a trace passing between pads of a 0.100"-pitch DIP, and +/-10% should not be a problem.

I suspect that the broadening of the lines is a result of the toner transfer step from the charge patterned photoconductor to the substrate. There are natural repulsions between the charged toner particles. The more toner that you put down, the more that this spread would mainifest itself. It might even be possible for the toner to spread or "bloom" on the substrate prior to the fusing step. Charge leakage in an optical photo conductor is also possible with an old worn out OPC, but the dark-state dielectric strength of these materials is pretty robust when within the normal service life. I agree that the spreading errors are insignificant in most cases, but may be in issue with closely spaced lines/pads. Dry film resist is less tolerant to conductor spacing than line width. The resolution of a typical (MG chemicals) resist itself is 80 micron lines with 200 micron spaces. If you are spreading 50 microns from each direction, this might explain why some folks had trouble with shorting on the tight-toleranced artwork used for some of these DIY projects. Like ebeam mentioned, it might be worth the $10 to get a pro print done at a graphics shop if there are tight tolences in the artwork. There was a shop in Cambridge that I used to use that had a 3600dpi printer for doing pre-press work. This will certainly save a lot of headaches associated with both overexposure and shorting.

Thanks for the micrographs.

I am more than happy to help with whatever I can give back to the community. I'm only starting to learn about actual electronic applications and design (as opposed to physics 101 type stuff). I'm sure I'll be hitting the group up with some more questions in the near future.

Regards,
Chris

 

[Tekay]

I use AGFA transparent film for inkjets, costy but works very good! Got it from a sceeningcompany that uses it for originals! Around $1.00 USD / A4
Printing the same layout on a Epson film is useless!

 

[StephenGiles]

Erm.......what the hell are you all on about? There is so much jargon on this thread......sorry!
Stephen