Electronics > Projects, Designs, and Technical Stuff

Toner Transfer

<< < (6/11) > >>

james_s:
The fact that I can go from design to finished board in under 30 minutes is reason enough to keep a the tools around to etch boards at home. Cheap PCBs from China are great, but when I want a one-off board NOW there is no substitute for etching it at home.

KL27x:

--- Quote ---Maybe your good enough to create a perfect board every time, but I know I'm not.
--- End quote ---
It has nothing to do with skill and everything to do with having done it hundreds of times and finally piecing together what causes the failures and stumbling on ways to take those failures out of the picture.

I had read about pre-etching, and the first couple times I tried it, my reaction was the same as most people. I didn't see the point. After trying it on a large batch of boards, I still wasn't convinced of any difference. It was only after I stopped doing it that I suspected an increase in error rate and that I noticed a difference in transfer quality. Namely, some of the failed non-etched boards had areas of fattened, uneven traces and even some grossly visible shorts. It is only on the area of the boards that I had "messed up" that the pre-etch made a difference. With the pre-etch, those areas that got too much heat were unaffected. I couldn't have known they should even BE different if I hadn't stopped doing it, AND I hadn't inadvertently overheated some of the boards that did not get the pre-etch. AND I happened to be using Pulsar, which doesn't have the chia pet mode of failure when the toner gets too hot.

Once I noticed this difference, I experimented with more heat on the pre-etched boards. Not just a little more heat. Wayyyy too much heat. I found the board fails before you get a significant degradation in the transfer. And now the failures due to too little heat? Those are history, too. It's a sure thing. So the problem... there's some scant info out there about pre-etching. But no one tells you that the benefit gained is that you get to crank the temp up thru the roof. This somehow keeps getting lost in translation. Maybe the first person who started doing it gained some benefit, but he did not even know what it was, because he never consciously tried to use more heat. Maybe his temps drifted up slowly over time from his initial starting point (learned without the holy combo of fiberless transfer medium and pre-etched board), but he was not aware of it. Who knows?

So.. I don't post this stuff to show off. I'm sharing information. It would be nice if every time I described what is basically the solution to the biggest problem with toner transfer that someone doesn't correct me and say it's unnecessary. This, unfortunately, is a losing battle. I'm just ensuring that the "pre etch myth" stays alive a little longer, so that people can continue telling others that the pre etch isn't necessary for another 20 years. If they take satisfaction in that, then I guess I did a good thing. And if one person every 20 years actually learns the difference and importance of the pre etch, and he tries unsuccessfully to share it, maybe the myth can stay alive forever.

"I do toner transfer with no pre-etch, and I get great results." Yeah. I believe it. I believe someone out there can strike out Barry Bonds 99 times out of 100. With the pre-etch and the Pulsar, this stretches the strike zone out to about 50 foot radius circle. So now I can strike out Barry Bonds all day long by setting my catcher up 30 feet away from the batters box. (I might give up a few first base steals, though, given the limitations of a human to catch errant throws :))

I'm also not for a second going to claim that I can out of the blue make a pcb in under an hour. Maybe when I was 20. Nowadays, you can watch the gears turn while I try to remember and piece together all the steps. There is a lot of standing around and drudging back and forth inside, to outside. Getting out the cutting board and the steel wool. Firing up the laminator. Waling round the house to fill the bucket with water. Getting a tank of air to pressure and hooking it to the tank. Alignment of the two sides with the drill holes in the board. It is just a complicated process.

Plus there are pre steps. Cutting up the transfer paper into more economical sizes and taping the squares onto printer papers in advance.

AE7OO:
Greetings,

KL27x, going back and rereading my post it is obvious that I made a boo-boo.

I was not taking about board CREATION so much as board DESIGN.  I thought it was implicit that I was talking about design as the last sentence  mentions me sending the board off to a PCB house and knowing I would receive a working board back(as long as the place made no errors of their own).

What I should have said is that I don't know about you, but I know for a fact that not every board I design works perfectly 100% of the time.  I may have to add or delete circuits, reroute parts of the board or change parts.  Creating boards myself allows me to respin a board without with that 5+ day delay for each version.

NOTE: I'm not going to try to create a BGA or a super fine pitch board  at home.  That nrf24L01P board  I did was a nightmare ( I did get two to work though).  Since then I've stuck with larger parts( about 98% of the time, that other 2% of the time I'll create the board and shoot it off the OSHPark and cross my fingers that I did not make a major mistake or forgot parts or circuits.  As a matter of fact just last night I totally spaced a simple STM32F030 board and forgot a LED circuit).

GB

cdev:
I totally hear you on the usefulness of short turn around. Thats the most useful aspect of it for me. I use the simplest etching solution I know of, hydrogen peroxide, salt and vinegar, and the crappiest free paper I can find, from junk mail I receive, I etch in a tupperware container I use only for etching, I just use a piece of insulation-coated stiff wire to hold the board by its edges, some IPA to use to clean the board of grease, a scouring pad, a flat board and an iron. Some blue 3M painters tape, And some nail polish remover. The beauty of it is its simplicity and the fact that I can etch boards while doing other stuff at my computer. I can agitate the solution every minute or so - sometimes I  just lift the plastic container up and drop it down every 30 seconds or so with my foot.. every few minutes and when I finish with a board and do another one I'll pour in a bit more salt to get it going again. Etch time is fairly short. At the end I pour the liquid into a plastic bottle I have and let the water evaporate away. The gunk from all the etchs I have done to date is represented by a layer of crystals in that bottle, at some point I will likely dispose of it at our local yearly hazardous waste disposal event.

 Its really easy to make small boards, its a very multitasking friendly activity, without much work required at all after the design is done and printed.

The drilling of holes for 'vias' or through hole parts is the most boring/annoying part. I would say it always takes less than a hour this way if you don't count either that (at the end) - its very time consuming if you have a lot of them, or the design process.

the resolution is likely better using other methods but doing it this way works fine for everything I've needed to do, with the caveat that the vias are the weak aspect of this method. So RF PCBs are non-optimal if you need lots of vias, this isnt the method for you. You'll find you need to design differently. It presents the least problems with RF passive circuitry, but active devices can be done with a little thought, just break your project up into pieces one board for each part, then use the entire back of those boards for ground and the front for signals and use multiple hand done vias around the parts to keep your grounds short. They sell rivets for this but I just use copper wire and side cutters.

KL27x:
AE700, ok. Thanks for the clarification. And for sharing your process and details of the resolution and success rate.


--- Quote ---That nrf24L01P board  I did was a nightmare ( I did get two to work though).
--- End quote ---

I suggest you may have found yourself in a less than optimal process, going by the info you posted. And maybe it's good enough for you. But with Pulsar, pre-etch, a laminator AND a 1200W heat gun, I do 8/8 mils on 9x6" panel with 99% success rate. Those boards that come out almost perfect but are unusable because of traces falling off or smearing together in an area of tight pitch, those problems are just gone.  You can get the transfer perfect, even all the way to the very edge of a piece of copper clad, every time. There's no prayer involved.

As for why? There are some things I produce with toner transfer. Dave did a recent video on "mod boards." I have had occasion to modify one of my production PCB's with such a board. There are only a couple of tiny components on the "mod board," and it gets soldered to another PCB. I can make it on super thin single sided copper clad with zero kerf. Just a panel packed with pads and components, which makes it easy to assemble and cut into individual pieces with just scissors. This takes off the issue of quantities and logistcs. How many will I need? I dunno, but when you do an assembly order, esp, the more you buy the cheaper it is. With a DIY board, I can just print off a bunch of panels onto the transfer paper, transfer and etch a few panels at a time, and populate and depanel as needed. If this approach is scrapped out of the blue, I don't have another bucket of PCBs to dump in a landfill.

I also tend to use toner transfer for one-off stuff like breakout boards, adapters, pogo pin interfaces, and stuff of that ilk. Stuff I us in test/production gear. Sometimes it only serves purpose of one project and another project will need a different board/tweak. That capability is here and it takes only a few hours.

Another use is when I am using a new part. I might toner transfer a test board just to verify my understanding of the datasheet. Just a small module to test that one component, before I design an entire PCB that includes it. The test board could be something super simple that needs to be used in conjunction with a solderless breadboard and/or other blocks to even be functional at all. If you do it yourself, you spend more time doing the work. But esp on these little rush, throw-away jobs, you have the benefit of never having an "engineering inquiry" that takes 3 days to resolve over email, simply because you rushed over the CAM job or have some simple mistake that would take only 20 seconds to physically bodge after the board is done.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod