[Top] [All Lists]

Re: [TowerTalk] How do you get a good looking trace when etching?

Subject: Re: [TowerTalk] How do you get a good looking trace when etching?
From: Jim Lux <>
Date: Sun, 03 Aug 2014 09:16:21 -0700
List-post: <">>
On 8/3/14, 8:38 AM, Roger (K8RI) on TT wrote:
Over the years I've tried etching boards with various methods. I have
never gotten a good looking trace. Whether the trace area is cleaned and
painted on, or photo sensitive, the edges are always irregular.  They
are not straight and taper to a thin edge. Just what you don't want.

Temp, chemicals (Typically Ferric Chloride IIRC), chemical dilution,
agitation: All according to directions with the same results.  I finally
gave up.

Kind of straying a bit, but since we're building LC tuning networks and antenna switches, and "tapered edges" are bad from breakdown standpoints..

There's a huge amount of difference between how commercial boards are made and how home boards are made. You may not be able to get "commercial fab" quality.

The etching process is a tricky one, and they adjust the chemistry and process to get a straight down edge, neither tapered out (like you have) or undercut. I'd venture that none of the commercial houses use a chloride etchant, the primary virtue of which is that it's "safe". Given the reek of ammonia at the last fab I visited, I'm going to guess that they use some ammonia complex in their etching. They also spray the etchant onto the panel, with filtering and reprocessing of the etchant (probably to insure that the chemistry *at the etch point* remains consistent).

I think they also use a multi step process commercially, where they put a photo resist on, expose it, then electro plate something ONTO the copper (tin?), then strip the resist, and etch the copper that isn't plated, using the other metal as the resist.

30 years ago, a friend and I were fooling with using nitric acid as an etchant when making nitrogen laser panels.. very fast, but there are some handling issues. But we got real nice edges with very simple resist.

These days, there are so many quick turn prototype board houses out there that it's almost not worth trying to make your own boards. After all, it's not the etching that is the hard part: it's doing the holes, plating through, etc.

Much easier to send a file to a fab and get your board back a couple days later.

The challenge, I think, is that for doing antenna switches and switched LC networks, is that the boards are physically large (esp if you want to handle high power), and the board houses charge quite a bit. Get much above 5-10 square inches and you're out of the 3 boards/$50 quick proto bucket.

For instance, Sunstone's Value Proto for a 6x8" board (I was laying out a big switched LC network) is about $450/3 pieces..

for a 45-64 square inch board 2 layer, no solder mask, no silk screen, Sunstone can do 2 boards for $262, 4 boards for 330, etc. That's for 1 day turnaround.


TowerTalk mailing list

<Prev in Thread] Current Thread [Next in Thread>