Believe it or not, the way you lay out your breadboards can make a huge difference in how well they work.  You can save yourself literally hundreds of hours, just by having a system and sticking with it.  Problems that used to plague you will magically disappear.  And I’m not talking about electrical issues like parasitic capacitance.  Don’t get me wrong: those things can happen, and you may be in for a few hours of troubleshooting when they do.  But once you discover a few simple rules and stick with them, you’ll realize the majority of your troubles were caused by simple, easily-avoidable mistakes.

These are the guidelines I’ve adopted over time.  Yours may be different.  That’s fine.  If what you’re doing works for you, don’t feel like you need to change anything.  If you’re just starting out, though, pay attention.  This will be on the test, the moment you find problems with your circuit.

Now, I’ll grant you that part of learning electronics is the exercise of troubleshooting.  But fate will hand you enough of these “opportunities” as it is.  You don’t need more. Would you rather spend your time troubleshooting, or admiring your final, working prototype and moving on to the next challenge?

 

1. Neatness Counts

This is rule number one. It’s a LOT harder to troubleshoot a rat’s nest than a neatly laid-out board. Many of the other rules are just corollaries of this one.

“But Kendall,” you may be saying, even though I can’t hear you through your screen.  “It doesn’t matter how you lay it out. If the circuit is the same, it will still work.”  That’s true.  If — IF! — your messy-looking circuit is indeed wired correctly, you’re right.  But on average, messy-looking circuits have more defects.  And it’s harder to find the mistakes in them.  When you finally find the offending part and it’s buried under other parts, you have to tear out other stuff just to get to it.  Will you remember where those components were connected?  Are you sure?  Great: even more opportunity for error.

Visibility is paramount. If you can’t clearly see where a component is connected, you’ll think it’s connected somewhere else.  3-D perspective can make a row of breadboard holes look like they’re off by one – like, the row that looks like it’s connected to pin 6, is really connected to pin 7.  Adding more components compounds the problem unless you — well, we’ll get to this one right off:

 

2. Keep Components Flat on the Board

Keep your components and wires flat on the board as much as possible.

I mentioned the 3-D perspective problem in the intro above.  When your components are standing high up off the board, overlapping and crisscrossing each other like a tiny insect jungle gym, this problem is exponentially worse.  To your brain, every different angle of viewing is interpreted as a different layout.  You dart your head around from side to side, trying to see where everything connects, trying to find parts that are hiding behind other parts.  Why fight it?

The secret is to work in two dimensions, not three. You reduce the complexity eight-fold.  The sun emerges from behind the clouds, and everything falls into place.  Suddenly your layout looks the same from any angle — one truth, one perspective, one circuit — and you realize that the land of breadboard nirvana is a very flat land indeed.

When I first started breadboarding, I used to dither for minutes choosing the best placement of each component.  I think I was trying to look several components ahead, and figure out how to do it with the fewest possible wires.  I didn’t want to clip any component leads.  What if I clipped a resistor to fit neatly across three holes, and then discovered later that I needed it to be longer?  I didn’t want to waste resistors, right?  So I left the leads as long as possible, and just stuck them wherever they needed to go.

Plus, with everything flat on the board, I’d have to use more of those really short jumper wires.  I only have so many of those.  And everyone knows that shorter traces are better than long ones, right?  More wires means more connections into the breadboard, too, with more chance of parasitic capacitance and all that…

Maybe I’m the only one that ever obsessed over this stuff, but somehow I don’t think so.  If any of this sounds familiar, then take my advice: don’t get hung up on that stuff.  In the world of DIY analog audio, where “megahertz” is a four-letter word, 99.9% of the problems in your layout will have nothing to do with parasitic capacitance or trace length.  99.9% of the time, it’s just you, plugging something into the wrong damn hole again.

Don’t be afraid to use jumper wires.

If you don’t have enough, go buy some 22-gauge solid wire and roll your own.  I have 20 rolls of the stuff, in all kinds of colors.  When I start getting low, or if I can’t sleep and need something to do, I’ll just sit and listen to the radio while I cut and strip wire, until I have enough.

Don’t be afraid to clip component leads.

Unless you’re breadboarding a precious 2N39 or something, don’t think twice about clipping leads to whatever length you need.  Resistors, capacitors, diodes, and general purpose transistors are all cheap.  So cheap, that you half expect to see them sold by the pound, outta those bulk-food-type bins you see at the grocery store.  (That would be cool. Radio Shack, are you listening?)  So get tons of them.  Get a lead-bending tool, too.  Count the number of holes you need to span, use the lead-bending tool to bend the component to that length, and chop off the ends, leaving about 0.25″ on each bent end.  Then just lay it in the board, nice and neat.  If you discover later that you made it too short or too long, you can usually re-bend it if you’re only off by one hole (0.1″).  If not, throw it back in its tiny little drawer or envelope, get a new one, and try again.  You’ll end up re-using the “wasted” one eventually.

But what if a part needs to cross over a chip or something?

Well, in that case, you have two options:

  1. Go ahead and clip it to the right length to lay flat on the chip, or
  2. Route around the chip.

I almost always do #2, even though it means using up precious board space.  If it’s just a plain wire, like between pin 2 and pin 6 of an LM555, I may grab the shortest wire possible and route over the chip.  If it’s a resistor, maybe, but not usually.  If it’s a capacitor or anything else, no way.  Run a flat wire to the nearest row that’s not connected to the chip, then another jumper straight across the gap, then another to the destination.  If that sounds like a lot of wires, just remember: it beats tearing your hair out troubleshooting.

And before you go through all my breadboard pics looking for how much of a hypocrite I am: yes, once in a great while, I will break this rule, when I’ve already used up all the available routes around the chip, and there are no better options.  But I feel really guilty about it later and can’t eat or sleep for days.

 

3. Double-Check Everything Before Applying Power

This one is so important, I moved it up to number three, and had to renumber all the rest after it.

After spending an hour or two laying it all out on the breadboard, I know you just want to hit that switch already and hear that puppy sing.  Well, slow down, cowboy.  Chances are, there’s at least one mistake in there and you don’t even know it.  “Well, dummy,” you say all sarcastically, “how can I know for sure, if I don’t turn it on and see what happens?”

Well, OK, that’s one way to find out.  Kind of like checking for leaks in a rocket’s fuel line by launching it to “see what happens.”  Sure, a rocket is way more expensive than an SSM2040 (though maybe not for long).  But there is another way.  A tedious, mind-numbingly boring way.  Trust me, though: even if you don’t like doing this, you’ll love the results.

While you’re laying out the circuit, get out a pencil and put little X’s through each component connection on the schematic as you make the connection.  Don’t cross out the components themselves, but cross out the connections between them.  After you’re done, and your hand is itching to turn the power on, go back and do it all over again.

What?

Start with one of your X’s somewhere on the schematic, check the breadboard, and make sure the connection is correct.  Don’t just glance at it, like “yeah, there it is, I know because I just put it there two minutes ago, duh.”  Really check it.  If it’s supposed to connect to +12v, is that rail you connected it to actually the +12v rail, or did you connect it to -12 or ground by mistake?  Did you connect to the right number pin, but on the wrong chip? Think.

If the connection really is correct, draw a little circle around your “X” and move on to the next connection.  Repeat until done.

Then fire it up.

I just started doing this recently after a particularly hellish multi-day troubleshooting session.  When I finally traced the problem to a mis-oriented JFET, I decided I’d had enough.  Every time I’d ever built anything on a breadboard, I ended up finding dumb mistakes eventually.  If mistakes were almost guaranteed to be lurking in there anyway, why not hunt them down before they have a chance to bite me?

And, by golly, it works.


4. Pick an Orientation and Stick With It

Are you one of those rare, clear-headed, well-rounded individuals that likes to work with the breadboard vertically, with the bus strips running neatly down the left and right?  Or are you one of those anti-social types lays them out horizontally, carelessly cutting across all common sense and order in an evil violation of electronic feng shui?  Well, either way, pick one and stick with it.

Just as moving from 3-D rat’s nests to 2-D flatness reduces complexity, staying with a single orientation simplifies layout even further.  It’s less work that the brain has to deal with, and electronics keeps it quite busy as it is, thank you.

Does it really make a difference whether you work vertically or horizontally?  I suppose not.  Horizontal seems to be more popular.  Sometimes it seems like I’m the only one doing it the right way.

 

5. Negative Rail on the Left, Positive on the Right

Like orientation, this is mostly a matter of taste.  Again, pick an arrangement and stick with it.  Personally, I use those breadboards with the double bus strips on each side.  Each side has a rail marked with red, and another one with blue.  I reserve the blue rail for “ground” (on both sides).  The red rail on the left carries the negative voltage, and the one on the right is positive.

This works out really well, because in most of the analog ICs I’ve used so far, the negative voltage pin is on the left side, and the positive on the right, when pin 1 is at the top of the chip on a vertical board.  I just have to run a couple of 0.4″ jumper wires from the far end of the pins’ rows to their voltage sources, and sometimes one to ground, and it’s all nice and neat

In most DIY audio, you’re going to be working with bipolar power.  If that’s not you, though, and you’re working with a single positive voltage and ground, you’re probably working on something boring.  You can come up with a different system.  But have one.

 

6. Start with the ICs, and Move Out from There

In a lot of DIY audio circuits these days — voltage-controlled oscillators, filters, amplifiers, phasers — the universe really does seem to revolve around the haughty analog ICs at the center. If your project’s schematic has more than a couple of those little triangle thingies, keep reading.

1. Lay down the ICs.

The first component I put in the breadboard is always the IC.  If there’s only one, I’ll try to put it in the center.  If there are two, I’ll leave plenty of space between them, and at the ends.

2. Connect them to power and ground.

Pretty simple.  I always use the holes furthest from the pins.  Then I connect up any pins that go through single components to ground or a power rail, like bypass caps and so forth.

3. Hook up pins that connect together.

Op amps use feedback, and that usually means connecting pins together on the same chip.  Sometimes it’s a straight connection.  Other times, a resistor and/or capacitor are in that feedback loop, or maybe more.  Regardless, I usually hook these up next.  If the loop contains more than one or two components, I’ll usually have to route to the “blank” areas above or below the IC, and then back.

4. Hook up anything that joins to the next IC.

I try not to work with more than one IC at a time.  It’s too easy to get confused about which IC I’m working on, and I’ll hook things up to pins on the wrong chip.

5. Hook up everything else.

This is where things get a little dicey.  By now, the rows next to the IC pins are pretty crowded, and the rows immediately above and below the IC are also full.  If you’re hooking up a particularly busy little chip, you may have to route some stuff pretty far out there.  But at least you got the easy stuff in already.

6. Last of all, hook up potentiometers, switches, and other “outside” components.

By leaving this to the end, it’s easier to put components and jumper wires in without the wires to the outside components getting in the way.

Yay, you’re done!  Now go double-check it all like we mentioned before.

 

7. Isolate AC Inputs and Outputs

This is where parasitic capacitance comes into play.  There’s always a little capacitance — a few picofarads — between adjacent rows on the board.  If you route your chipamp’s output next to a row that’s connected to ground, guess what?  Most of that glorious amplified sound is probably draining out of through that picofarad “connection.”  The same kinds of things can happen if you route something carrying a high-current AC signal next door to an op amp input, or to something that feeds that input: unwanted feedback.  “High current” here usually means an end output stage pushing 1 volt or more through a speaker or other low-impedance connection to ground.  So, things like a chipamp output, or a 5v clock signal feeding CMOS chips or something.  With low-level input signals, like from a microphone or an unamplified oscillator, you’re probably fine.

Parasitic capacitance can be tricky with CMOS logic circuits.  Things may still kind of work. When I breadboarded a Baby 10 sequencer for the first time, the CD4017 chip was really finicky about advancing from one output to the next.  Sometimes it would do it, but usually it wouldn’t.  If I increased the signal voltage from 9v to 12v, it worked better.  I tore my hair out for a couple days on that one.  Turns out instead of connecting the LM555 clock output directly to the 4017 clock input pin, I’d connected it to the row right next door.  Parasitic capacitance let through just barely enough signal to trigger once in a while at 9v and more often at 12v.  Lesson learned.  (By the way, if I’d followed Rule #3 above, this would not have happened.)

 

8. Mark Interesting Places with Flags

Like this one:

Just wrap little adhesive labels around the wires and you’re set.  I use these to mark inputs and outputs.  It’s especially handy for remembering where to hook everything back up after I’ve disconnected the board.  Plus it makes it easier to show other people in a photo for troubleshooting help.

 

Got more?  Post a comment and share your techniques.

Comments

Leave a Reply

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>