Solutions to shorts/issues with wired components

[newvoodoo]

It seems if I have a circuit that has few or no wiring required, it works perfectly when I first power it on..and when it has a number of components connected to the board via wires (potentiometers, switches, jacks)-the more, the more likely I am to have a problem.
Ive tried solid v strnded wire and seen little improvement in solid, certainly not enough to overcome the difficulties inherent (less malleability, takingm ore space), and i've tried some other things.
Now ive taken to double wiring important things like in/out jack connections (meaning I'll use two wires intertwined for one connection) and that seems to help a lot.
Any other suggestions?
Would a more conductive solder work like one with a lot of copper or silver?

I should point out i've been DIY'ing electronics for a good decade or so-so suggestions like "learn to solder" will not be helpful
Im just frustrated enough to hope someone else has a good idea since I've been doing some things lately like multi-board pedals and synths and so forth..I built MFOS's analog guitar synth for example (used about 50 feet of wire and has roughly 30 wired pots and 10 or so swithces, few jacks)..and its REALLY frustrating to have the same problem turn up over and over without a solution.

thank you for any time anyone takes in reading this and providing any suggestions, I wwelcome them all-its certainly possible im overlooking something that a total newcomer wouldn't or somesuch.  Im DEFINATELY looking for advice and willing to take any at all.

[R.G.]

Actually, you have discovered a fundamental law of nature.  

Everything has different qualities, and you pay for the good stuff by dealing with the bad stuff. Wires give you mechanical flexibility, independence (you can remove/replace as little as one wire, not a whole PCB) and anywhere-ness, but require you to pay for that in dealing with organizing the wire routing, making two additional connections per wire, and dealing with the electrical side effects of crosstalk and so on.

Preferring to use hard connections, like on a PCB, gets you simple, easy connections, as they're already in the right places every time, fewer joints to make; but it costs you in that you can't move things to different places and there's no mechanical flexing allowed.

I personally hate cutting, stripping and soldering wires. But there are times when nothing else will do. I once designed all of a tube amp on one PCB - and then put every control on flying wires to get immunity to vibration and the ability to replace a pot or switch without having to pull the PCB out.

I recently completed a PCB that replaces 100% of the insides of a Thomas Vox Beatle preamp. It has 76 wires coming off the PCB. But those 76 wires replace 87 wires coming off the original board, and they are place on the PCB where they don't go into a bundle here, there, and everywhere, and only act as a flexible "hinge" to be able to flex the board up for servicing. Sometimes wires can be dispensed with, and sometimes they're the only thing to do. The trick is to know and use them for what they're good for.

It's a good practice in all design work to identify what you get and what you paid in non-monetary terms.

[amptramp]

I always run into wiring problems with things that go off-board and sometimes with things that go on the board.  To answer one question, the standard 37% lead / 63% tin solder is hard to beat.  If you are using one of the lead-free solders, you will have more trouble getting good joints.  More conductivity would not really matter.

If you have flying leads coming off a board, you may have problems with fatigue failure where the leads are flexed.  You may be best off allowing more lead length and holding the lead to the board with zip-ties so the flex does not come at the delicate solder joint.

I have had good luck with AWG30 wire wrap wire for interconnects and it has the advantage of low capacitance to anything else.  The Kynar insulation is quite resistant to abrasion which is not the case with Teflon or PVC.

Consider using board connectors rather than wiring from each individual point on the board.  You can remove the board without disturbing the rat's nest of wiring or having to physically remove pots etc.  from the box.

I still prefer to have jacks connected by flying leads to avoid putting stress on the board when inserting or removing the plugs.  Similarly, the last bit of torque used to install a pot may crack a trace on a board and if you put pots on a board, you have to get the positioning and alignment just right to match holes in the box.  Sometimes it works well but sometimes it just isn't worth the effort.

I admire the people who design complicated circuitry and put it in a 1590A.  I admire it - but I don't do it.  I prefer to leave plenty of room around any board to do the wiring and to make sure I have enough lead length to remove things from the box without overstressing the wiring.  Even in spacecraft I worked on where weight was the dominant specification, a service loop was provided so stuff could be removed without breaking a wiring harness.

[R.G.]

I always run into wiring problems with things that go off-board and sometimes with things that go on the board.  To answer one question, the standard 37% lead / 63% tin solder is hard to beat.  If you are using one of the lead-free solders, you will have more trouble getting good joints.  More conductivity would not really matter.

Good post in general. Just to add some bits in support:

If you have flying leads coming off a board, you may have problems with fatigue failure where the leads are flexed.  You may be best off allowing more lead length and holding the lead to the board with zip-ties so the flex does not come at the delicate solder joint.

Fatigue cracks at the board or component joint are the biggest issue with wires that get any flex at all. Breaks are almost never in the middle, undisturbed part of the wire. Breaks happen where there is some boundary to concentrate the stress.

The worst of these is often at the point where the insulation was cut to strip it away. Any wire stripping technique with a mechanical-cutting blade generally nicks or compresses the copper wire right at the end of the insulation. The insulation makes the unstripped wire stiffer, so it tends to bend the nicked place more than the rest of the wire, and so the stress fractures grow from the nick. I have known some instances where a high-end buyer would require all thermal wire stripping to avoid exactly this.

Even then, the insulation concentrates stress there.

A sneakier way of getting this is when solder wicks up a stranded wire under the insulation. This reinforces the nicked place, but then you get stress concentration at the end of the solder wicking. That can be a tough one to debug.

Ways around stress fractures include drilling an extra hole in the board and running the free wire down through the board again after the stripped end is soldered in. The insulation in the extra hole in the board buffers the wire, and the hole prevents movement. Solder wicking is generally kept to the zone before the hole. Another one is to squirt a blob of non-corrosive RTV around the soldered pad and insulation on wires. This mechanically keeps the wire from bending so much at the nick/solder, performing much like the extra hole.

Solid wire above about 24Ga is a disaster if there is any nick at all in the wire. It breaks at the nick. The Vox preamp replacement board was conceived and executed because  Thomas Organ used solid core hookup wire in about 24ga, and ran wires all over the boards and then laced the mess into a bundle. Service on the boards is a nightmare of fixing something, then having something else quit because a wire popped off.

I have had good luck with AWG30 wire wrap wire for interconnects and it has the advantage of low capacitance to anything else.  The Kynar insulation is quite resistant to abrasion which is not the case with Teflon or PVC.

30AWG and Kynar works because it's very flexible indeed, and wire wrap strippers tend to be specially constructed to not nick the wire.

Consider using board connectors rather than wiring from each individual point on the board.  You can remove the board without disturbing the rat's nest of wiring or having to physically remove pots etc.  from the box.

Yep. If you MUST run wires from a board, make all of the bundling complexity be IN the PCB, and run all the wires in coherent bundles, not in rats' nests.

I still prefer to have jacks connected by flying leads to avoid putting stress on the board when inserting or removing the plugs.  Similarly, the last bit of torque used to install a pot may crack a trace on a board and if you put pots on a board, you have to get the positioning and alignment just right to match holes in the box.  Sometimes it works well but sometimes it just isn't worth the effort.

Exactly the reasoning I used in the Workhorse amps, aided by a grizzled tech showing me how to work on a PCB Marshall: remove 18 knobs, remove 18 nuts/lockwashers/washers, remove 10 board mounting screws, remove the PCB from the chassis. NOW you can start debugging. And when you're done, put them all back in the reverse order, having not dropped one under the bench during the repair. Every pot on the Workhorse - and the Vox replacement, for that matter - has exactly three wires from the PCB to the pot, and no where else.

I admire the people who design complicated circuitry and put it in a 1590A.  I admire it - but I don't do it.  I prefer to leave plenty of room around any board to do the wiring and to make sure I have enough lead length to remove things from the box without overstressing the wiring.  Even in spacecraft I worked on where weight was the dominant specification, a service loop was provided so stuff could be removed without breaking a wiring harness.

Amen. I admire sword swallowers and people who lie on beds of nails - but, uh-uh, not me.

[tubegeek]

Breaks are almost never in the middle, undisturbed part of the wire.

I had to fix a mini mic cable yesterday that was one of the very, very rare cases of a break in the middle of a cable.

As my friend Ken, sax player, described the events, the cable from his sax-bell-mounted mic to the XLR-adapter belt-pack got snagged somehow and he yanked it violently. I tried testing from each end but there was no hope for it - the cable was really, truly broken somewhere between the two ends, and there was no obvious spot where the damage looked to have been.

So I soldered his min-XLRs onto a new piece of cable. It does happen but I think this was a first for me. He says it'll result in a Thank You in the booklet for the CD he's recording, so I got that going for me, which is nice.

Big hitter, the Lama.

[karbomusic]

got snagged somehow and he yanked it violently

Which would make it nearly impossible to trust anyway. There could have been multiple breaks anywhere along the wire with only a few strands still intact (due to the violent yanking) making it look like it was fixed only to break again very soon. On a side note, band members used to hate me when I would find an intermittently working cable being used on stage and I would immediately cut the connector off. Because, if it is intermittent today, it's going to tomorrow and the next day and always at the very worst time.

[PRR]

> double wiring important things .... that seems to help a lot.

Very strange. If I make one bad connection, it's sure that I can make another bad connection. That double-up helps you is A Clue, but I can't figure it out.

> Would a more conductive solder work

Your runs are 99% copper wire and 1% solder. Solder aint as good as copper but it isn't a "bad" conductor by any measure. Probably 1,000X better than any stomp-joint needs to be. Sexy-solder is usually about metallurgical issues (gold-dope solder when we had gold-plate leads).

The "un-conductive" part of solder joints is tarnished metal. Tarnish control is key to making 99.99999% good solder joints. However you say you have been soldering for a decade. So I am sure you know to scrape un-fresh lugs and leads, use appropriate flux, pre-tin lugs and leads, melt the whole joint, see that it "wets", and let it harden without movement. Look at your work, and study your failures until you see the flaw.

[GGBB]

Another one is to squirt a blob of non-corrosive RTV around the soldered pad and insulation on wires. This mechanically keeps the wire from bending so much at the nick/solder, performing much like the extra hole.

My off board wiring failures always happen during either assembly/dis-assembly or out-of-box testing/handling and always at the wire-to-board joint.  Since I've started using drops of hot glue at the joint, the problem has pretty much disappeared.  At the component end (pot, switch, jack) I use heatshrink for strain relief.  I've had the same issues with a few commercial pedals and took the time to apply glue in those as well.