The PCB Part 4: Drilling, Populating, and Soldering

I finally got around to finishing the PCB last night.  I didn't have any drill bits near small enough so I used some Amazon points to order this micro drill set from Amazon which included a swivel head pin vice (the "handle").  It's a pretty nice set but it only took me two holes to realize I didn't want to do the rest by hand.

It would've been best to use a small drill press, which I don't have, so I figured I'd just use a rotary tool.  I have the full Dremel kit (minus the drill press base, of course) including the pen attachment but that still seemed too bulky.  I bought one of those cheapo 80 piece rotary tool kits from Harbor Freight quite some time back.  It's nice and small, which would make it easier to keep the bit from wandering, so I thought I'd give it a try.

I planned to wrap tape around the 1mm drill bit from the Amazon kit to fit it into a collet but, low-and-behold, there is a 1mm collet as well as a 1mm drill bit in the HF kit!  This made short work of the remaining holes though it did have a tendency to walk slightly.

After drilling was complete, I used the real Dremel with the side cutting bit and the small, plastic router base to cut the PCB to shape.  This worked fairly well, so long as I went slow and kept a firm grip on the Dremel.  I used a hand file afterwards to clean up the edges.

Then, on to the soldering.  I discovered two areas following the etching process where the traces had breaks.  I addressed those while soldering but using a small jumper wire.  I also decided to "coat" the copper traces with solder.  My reasoning was to reinforce them, in case any others were marginal, as well as to provide a little corrosion resistance.  It looks ugly as hell but it works.

For the wired leads, I went with standard Yamaha colors (at least for circa 1980): Black = Ground, Blue = Indicator, Yellow = Brake.

The LEDs were purchased online from superbrightleds.com.  Their pricing was reasonable and they were shipped extremely quickly.  You can find cheaper LEDs from Asia on eBay but I was quite pleased with my order from these guys.  As an added bonus, the box they shipped in, although over-sized for the 50 LEDs I order, ended up being a perfect size to carry around the various parts of the light (including the housing pattern/form) while I work on them.

Each section of the light (indicator and brake) consists of three parallel strings of four LEDs and one resistor.  I calculated the required resistor for each to be 260 Ohm but, with 270 Ohm being the closest standard size, I went with that.  (Okay, okay... I didn't exactly calculate the size at first.  I used an online calculator but I did go back and confirm it mathematically, just to make sure I remembered how to do so.)  I also found I only had five of the six required 270 Ohm resistors so I had to put two 560 Ohm resistors in parallel to get it back down to the required resistance.  Good ol' Ohm's Law!

And best of all, I got the circuit correct on the first attempt!! Alright, not exactly the first attempt.  This actually was version 1.1 where I switched from standard 5mm "dome" LEDs to the four conductor flat mount type.  In doing so, I did find that I had placed one of the LEDs backwards in the circuit which was remedied in this version.  Either way, we have light!!

Indicator/Tail Lamp	Brake Light	Both Circuits

Lessons learned/things I'd do different next time:

• Using overhead transparencies instead of photo paper was definitely the way to go.
• I'd make the traces thicker as well as the component pads.  This would help in the etching phase but also in the drilling and soldering step.
• A drill press, though not required, would've made drilling so many holes a little easier.

The PCB Part 3: Etching

I finally did it, I moved on to the etching process.  As mentioned previously, I'm following a combination of various methods posted around the Internet.  (Standard warning: do your own research, don't take my word on any of this.)

The setup.  Since working with chemicals, in particular acid, scares me, I probably over did the safety gear.  I wore goggles, an apron, respirator (with P100 level filters), vinyl gloves, and a face shield.  I'm sure my neighbors thought I was cooking meth.  I also kept plenty of water nearby for rinsing and emergency flushing.


I went with transfers #4 on one side of the board and #6 on the other.  I used 3 parts H2O2 to 2 parts HCl as suggested in most of the posts I read.  (The zip tie on the board provided a convenient, non-metallic way to handle it.)







It did take longer to etch than I expected.  I think it was in the solution for almost 20 minutes whereas I thought it'd only require a couple of minutes based on what I was reading.  Granted, most of those examples used a much smaller board and I'm not certain of the copper thickness they used.  In addition, it was a bit chilly this morning.  I think it was barely 60 degrees in the garage so I'm sure that had an affect on the required time as well.

I should also note, I had transfer #4 facing down in the solution for most of the process.  That side seemed to etch about twice as fast as #6 which, I'm sure has to do with the Oxygen and other gases being released as part of the process.

After 10 minutes the copper is finally beginning to disappear on #4.












At the 15 minute mark #4 is done but the side with #6 looks like the other five minutes prior.  It was at this point that I began putting #6 facing down in the solution.

 


















20 minutes since beginning, both sides are finally done.

 















 
Here it is after rinsing the board thoroughly with water and storing the solution in a plastic container.  Apparently it is reusable provided you re-oxygenate it.



 





And, finally, both sides of the board after the toner was cleaned off with Acetone.  Side #4 didn't come out all that great but that was expected.  I think it was made even worse due to over etching since it was exposed so long after it was finished while waiting on side #6.

I was, however, very pleased with side #6. Again, it wasn't perfect but there were only two traces that had breaks and I can fix that when soldering.  Also, I should've been a little more diligent while cleaning off toner from previous attempts and you can see the remnants in the etch but that too can be touched up.  Regardless, it will be good enough for me to test my circuit layout and, if I got it correct the first time (doubtful), it may even be the board I use in the final tail light.  We shall see....

The PCB Part 2b: Transferring... again.

I decided to attack transferring the PCB again. Following the unsuccessful attempts the last time around, I made a few changes in my approach.  

First of all, I was more diligent about getting coverage and heat with the iron.  The second, I read a few additional methods that mentioned preheating the copper board first prior to applying the toner image.  And finally, I came across another site that mentioned using overhead transparency film in place of the photo paper so I added that to the mix as well.

Here's the first attempt (#4 overall) with the overhead transparency.  It wasn't perfect but is the best result so far.  I think I might be able to repair some of the incomplete tracks with a Sharpie before attempting to etch.  Another positive feature of using the transparency was, if you are careful, you can hold one end of the film and gently lift a corner to see if it is transferring adequately.

Still, I wanted to make it work with the photo paper, so on to attempt #5.  This was the worst yet.  I preheated the board as with the transparency, was very meticulous about coverage with the iron, and it hardly transferred at all.  I give up with the photo paper.






Since the copper board is double sided and my design is only a single layer, I intend to etch circuit on both sides.  That way, I can choose the best one and just sand off the other side. As I did #5 on the back of the same board of #4, I just cleaned off #5 using some Acetone and re-scuffed it with the Scotch-Brite to get ready for...

Attempt #6, back to the overhead transparency.  This is the best yet.  Again, it's still not 100% perfect but I think it's good enough to use to test the etching process. 

The PCB Part 2: Transferring the Layout and Etching

As mentioned previously, I'm using the toner transfer method to create my PCB.  I've read a lot of methods that used Inkjet printer photo paper as well as some people who have had success using glossy magazine pages, both printed using a laser printer.

I decided to try both since I already had the photo paper but, in the future it would be nice if the magazine pages work since I have plenty of those around.

	The bits....
	The board trimmed to size, scuffed with Scotch-Brite and the iron ready to go.
	Attempt #1 with the photo paper, not much of the layout transferred.
	Attempt #2 with the magazine paper.  I believe this piece came from a catalog I received in the mail.  A little more transfer but again, not complete.
	Attempt #3, back to the photo paper.  This one seemed more promising as I actually had to soak the photo paper to remove it.  But, after doing so, it was better but it still wasn't a complete transfer.

I'm too drunk and/or tired to mess with this any more tonight.  I have a couple of theories about what may be wrong (other than the obvious).  First, several of the articles refer to setting the laser printer to use maximum/darkest toner setting.  The one I'm using doesn't allow that option so I'm stuck with the standard amount.  Second, and most likely, I'm not being patient enough when heating the paper with the iron.  I'm probably moving the iron too soon and not being careful enough with overlaps which may be leading to the gaps I'm seeing in the transfer.

I'll try again at a later date.

The PCB Part 1: Layout

Along with preparing the form for casting, I've been working on the electronic side.  As mentioned, I plan to go with LEDs.  Rather than using a typical project board to lay them out, which would require a much more soldering as well as cutting/stripping way too many jumper wires, I figure I'd go for a custom PCB.  

Since this is truly a one-off PCB, it doesn't make sense to send it out to be made.  Even one of the discount makers, with the required order minimums, isn't cost effective.  

Plus, I've been wanting to try to etch my own PCBs for a while using the toner transfer method. I'm following advice from several methods from the Internet but I'm sticking closest to the write up on the Robot Platform website.

The first step is to create the layout.  I came across this really neat layout program called Copper Connection written by David Cook.  It is an excellent little program!  It doesn't have the advanced features  (such as auto-routing, circuit analysis, etc...) that some other packages may have but it is exceptionally simple to use.  And for this project, it was perfect!  I highly recommend it, check it out at Mr. Cook's website Robot Room.

After a few minutes to get up to speed, and one revision change after deciding to use a different type of LED, I was able to layout This circuit will include the tail lamp (i.e. the light that is on when the headlight is on) and the stop lamp (aka brake light).  I wanted to incorporate turn signals as well but felt that would be too complicated for a single-sided PCB.  

Here's a quick shot of the layout showing the top silk screen (which won't be used in the final PCB) and the bottom copper layer.  

Keep in mind, the bottom layer needs to be backwards because it will be transferred to the actual PCB.  One of the great features of Copper Connection is that it allows you to layout a component as a stack on all layers at once.  Then, when it comes time to print the PCB, depending on your selected transfer method, it flips the appropriate layers as needed.  This is especially useful after having a few beers, when your spacial orientation ability is not where it should be.