A simple 6502 SBC. Request for comments

[GARTHWILSON]

Dr Jefyll wrote:

But a fat VCC trace can actually be a disadvantage in certain troubleshooting situations, such as when a chip or a bypass cap has shorted. You're better off to have skinny VCC traces because their resistance is high enough that there'll be a measurable voltage drop across the trace that feeds the shorted component. (Have been there and done this IRL -- a tidy solution to an otherwise nasty situation.)

I made myself a short-finder for this purpose, that puts 2A pulses through, at less than 2% duty cycle, and I can use the oscilloscope on the 5mV/div vertical gain setting to probe around the board and track down the short, even between planes (which usually means where there's a shorted capacitor that goes from one plane to another, but you can even detect where a short is in the board itself!). Here's the sheet in my 3x5" workbench quick-reference guide:

The op amp is just a cheap LM358 8-pin dual op amp.

[BigEd]

Pulsing the current- I like that. Would an infrared camera help at all?

[GARTHWILSON]

BigEd wrote:

Pulsing the current- I like that. Would an infrared camera help at all?

An IR camera might help if you don't pulse the current but limit limit the current to the highest level the traces can safely handle. A 2A pulse for less than 2% duty cycle is less than 40mA average which will only raise a .008" trace's temperature by about 1.5°C (and take a long time to do it). If the traces are underneath components though, or on an inner layer, it will be extra difficult. For a short between planes (which hopefully will be at a component), depending on the component, the current may have to be awfully high to find the short with an IR camera.

[BigDumbDinosaur]

I should note that when building the first copy of a new design, I meter out the bare board before populating it to catch the shorts that accidentally got designed in the layout. If the metering out process seems to be okay, I attach the power connector, apply power and assuming the power supply doesn't fault out ("crowbar" in railroad electric traction parlance—not the technically correct usage), I check all points that are supposed to be powered for the presence of power.

With the above checks handled, I populate the board, periodically checking for a low resistance or short between Vcc and ground, especially after installing an electrolytic (POC V1.0 had a reversed electrolytic error that I caught during assembly-stage testing). By the time the board is fully populated I'm reasonably confident that I'm not going to power up into a catastrophic fault. That, of course, doesn't mean that a defective part couldn't go bang, but in practice, that has very seldom happened. In fact, it's literally been years since I last got a new IC that was defective.

The above process may sound like something a Nervous Nellie would do, but in practice it has saved the day more than once. Burned-up carbon comp resistors don't smell very good and I especially loathe having to clean up after an improperly-installed electrolytic blows following power application—had that happen in a 100 watt vacuum tube PA amp and, boy, did it make a mess of things.

Something that could be done when first-time powering a gadget is to insert a low wattage incandescent lamp in series with Vcc to act as a fault current limiter. I'm "braver" than that and depend on the power supply to automatically shut down if a fault occurs. That may not save parts but it will in most cases prevent PCB damage so a post mortem can be conducted.