Tuesday, August 2, 2011

The Spiral Challenge

I have been making a lot of Homebrew PCB lately, using the Toner Transfer Method. Must of the boards are prototypes, which will be later be fabricated by one of the online PCB suppliers. I have been feeling very good about the Homebrew Process and have created several small boards with high resolution for very dense SMD projects. The highest resolution projects are 6 mil traces with 10 mil separations between traces or ground plane. My more typical project are created with 15 mil traces and 12 mil separations. Some of the details of these projects have been posted in my previous blogs (note: the selection will include a copy of this post).

The Spiral Challenge
Several years ago, when I first started Homebrewing PCB via the Toner Transfer Method, my resolutions and yield - was Horrible!

I saw the Homebrew_PCBs Yahoo Group Double Spiral Challenge web page, that provided the art work for process challenge, with several standard width traces and spacing. The image is used to produce two concentric PCB trace spirals, where the acceptable resistance between and end-to-end specifications are given.

In my early days of trying to make PCB, I could only hope for something that actually worked, regardless of the asthetics or (then low) resolution.

Now that I can produce High Resolution Homebrew Printed Circuit Boards almost 100% of the time, I decided to take the Yahoo Group Double Spiral Challenge. I selected the (mid range) "spiral-10mil" Challenge as a starting point. It is 10 mil traces with 10 mil spacing. A quick calculation, suggests that each spiral is about 200 inches (16 feet) long.

2 inch radius / (10 mil trace + 10 mil space) / 2 spirals * 4 inch average turn length ) = 200 inches

Currently, I use a Brother 1200dpi Laser printer (I know, most Toner Transfer users suggest it can not be used). A modified GBC Laminator and Pulsar Transfer Paper.  See previous posts and my Salmoncon Take-Away-Notes.

I recently received a new stock of Double Sided PCB Material and wanted to try this new material.

My first attempt to Transfer an Toner Image, was flawed, a small hair or fuzz was trapped between the transfer paper and the PCB material. That is no real problem, I just re-scrub, reprint and to try again.

My Initial Results
On the second attempt, the Transferred Image looked promising, but the process  had it's share of problems, I must have been just too anxious for the results.

The first thing that I forgot in the process, was to tape over the back side, to avoid unnecessarily etching of the extra copper. The process took much more time and Ferric Chloride than I expected.  I thought this could be a problem.

From my previous Homebrew PCB build process, I know that the faster the etch and the smaller the copper area to be etched, creates the best results. This etch was taking to long!

The results, . . . . FAILURE !

The flaw is on the forth trace
above the ink spot
Resistance between traces implied there were NO shorts. But, end-to-end of each spiral was open. Under close Microscope inspection (and a with considerable amount of time), one spiral was found to be open in one spot, and the other was open in three spots. The one spiral is open on the forth trace, just above the ink dot on the photo.

All of the open flaws appear to be the results of etching through toner pin holes or single point under cutting. The opens did not appear to be the results of general narrowing of the traces. Each opens spot was shorter than the traces width themselves.

I will take the Challenge again, maybe even at a high resolution, but I will pay much more attentions to the details.

For now, I have a nice new, . . coaster.

Update: see http://wa0uwh.blogspot.com/search/label/Spiral


No comments:

Post a Comment