Minor question of PCB building philosophy

[G. Hoffman]

So, I'm making this MIDI loop switcher.  The way I am doing it, I have three boards - one for the Microcontroller and the Logic latches that do the actual thinking part, which I had made because I couldn't find a way to do it without at least a dozen jumpers, so I wanted a double sided board.  That one is fine.  Then there is a board for the 7-segment display and it's logic, which didn't fit on the main board, plus I wanted it on the other side of the box.  No problems there.

My issue is with the relay/switch board.  (4 relays for audio switching, and either a 4066 or 4016 for my amp/speaker cabinet options, which have their own relays.)  I made a mistake when designing it, and reversed the direction of the coils.  So, for each relay, I have one trace which is wrong, and I need to reverse the direction of the protection diodes. 

The diodes are easy, but the question is, do I make a new board for 4 short traces (about .1" length)?  I've already done the design change (a trivial change, other than the fact it doesn't work without it), and I could make the board in about an hour, maybe an and hour and a half.  OR, do I just cut the offending traces, and use my conductive pen to make the right trace?  Cutting the traces and all would probably be 15-20 minutes, so from a time point of view, the answer is pretty obvious, but I'm not sure how big of a deal it is quality wise.

By the time I read any answers, I'll have made the new board, but I'm curious how other people would handle it.  My reasoning is: this is for the coil of the relay, so it has a (relatively) high current, so I want to have a good real trace.  Or am I just making more work for myself than I need to?


Gabriel

[JKowalski]

I wouldn't make a new board. Cut the traces as neatly as possible then bridge the new trace with a piece of solid core wire.

[G. Hoffman]

I wouldn't make a new board. Cut the traces as neatly as possible then bridge the new trace with a piece of solid core wire.

You're probably right, but the new board is already made now. 

I have this problem.  I build guitar for a living, and if you are going to do that you need to learn to sweat every single, tiny, invisible, totally irrelevant detail.  Customers expect it.  It gets to be a way of thinking, and it's kind of hard to let go of, at least for me.

It took me longer than I wanted, though, as I had a bad toner transfer, and I was out of gloves so I couldn't use the etching method I like to use. 


Gabriel

[Processaurus]

I wouldn't make a new board. Cut the traces as neatly as possible then bridge the new trace with a piece of solid core wire.

Yes, a real prototype has at least a few awkward trace cuts and mods, I appreciate commercial stuff with ratty mods and components stuck on the trace side, like on old EH pedals.  It makes me think they were willing to sacrifice good looks and professionalism for good sound.

It's pro to try to do those kind of mods neatly though, tiny solidcore teflon wire is good for getting into tight areas, and surfacemount resistors and caps can be fairly discrete, like if you scrape the solder mask off a trace, cut it, tin the two sides, and stick the SMT component over the gap.  Drilling extra holes in the PCB for new wires is fair game too.  The art of the harmonious kludge...

[head_spaz]

Beauty is a perception thing, it exists only in the eyes of the beholders.
Physics isn't as picky about appearances, take a look around if you don't believe me. Lots of dead-fugly everywhere you look... but life goes on unnoticed.

In the category of stompboxes, function outranks form. Creativity and asthetics is greatly appreciated, but not absolutely essential. Electronics is about physics, appearance is about art.
The bottom line is to make it function.

One could argue that there's nothing glamorous about perfboard/vero... it's 100% kluge from the beginning. But the method works, and once the lid goes on a multitude of sins become concealed, yet it's still a stompbox.
Beauty is only skin deep, and stompboxes have skin made of opaque aluminium.

But like G. Hoffman, I'm a tad bit compulsive myself. Before I take on a project, I have to convince myself that I'm capable of doing it. And any job worth doing is worth doing right. Sometimes this re-quires re-doing things until they're re-righted.

[G. Hoffman]

Yeah, it turns out I didn't catch all the mistakes.

So I cut a few traces, and made added a few jumper wires.  If I ever have to make another one of these, I'll have to work the layout a bit, because one of the things I missed goes through a pretty tight channel!

Here are the boards in question, by the way.  The biggest of them was made by Express PCB, and I have to say that soldering (and unsoldering) plated through holes could get addictive!  I made the other two.





The replacement relay/switch board, before I found my mistakes:



Having just finished R.G.'s book, I'm not wild about the layout of this, but I don't really feel like starting over from scratch, so I guess it will do for now. 

{click for larger versions - quite large}


Gabriel

[G. Hoffman]

OK, looking at those full-sized versions of those images sure does show off some blemishes in the traces!  Wow.  Don't look too close at the solder joints, please!

Then again, this may be a better way than my normal magnifying glass to check every thing for problems.   


Gabriel

[JKowalski]

Oh, well if it was a production model then I would probably do the same and make a new board. I'm sure I would be self conscious enough about the product to force myself to do it. But for personal projects, I couldn't care less. I will settle for messiness because I know that if anything went wrong it wouldn't matter, because I have the capabilities to fix it. A bit of sloppiness can be excused for guaranteed warranty, so to speak.