A member of my Hardware Academy was about to order the first boards for his power management design, which powers and monitors three onboard cameras.
The schematic was finished, it looked clean, and the next step was paying for fabrication.
So before he placed the order, he brought it in for a design review.
One of our engineers worked through it and stopped at the current-sensing resistors.
They were wired in parallel with the power supplies instead of in series with the load.
On a schematic, that difference is just a couple of connection points, and it looks equally tidy either way.
On the board, a 0.1 ohm resistor straight across a 5V rail is a dead short.
The engineer didn't soften it: "It is short circuiting your supplies! This spells instant death to your regulators, you're drawing 5V / 0.1 Ohm = 50 Amps!"
The first time that board powered up, the regulators would have destroyed themselves before he ever measured a thing.
The review turned up two more problems.
A swapped net label had one camera's shutdown signal wired to the other camera's regulator.
On its own, that error would have powered up fine, then misbehaved in ways that are miserable to debug.
A battery subcircuit was missing entirely.
So he fixed all three right there on the screen: "Thanks for the catch on the current sensors and net label for the shutdown signal. I have fixed those."
The fixes cost him nothing, because no board existed yet.
If he'd skipped the review, that first prototype run would have arrived, died at power-up, and taken the regulators with it.
Then he'd be paying for a redesign and new boards, plus weeks of lost time, all to learn what one review caught in a single pass.
A messy schematic gets scrutiny, and a tidy one gets trusted. That's what makes an error like this so easy to miss.
A resistor in parallel looks just as deliberate as a resistor in series, and it takes someone tracing the actual current path to tell them apart.
You can't do that on your own design, and it has nothing to do with being careless.
You drew it, so your eyes fill in what you meant instead of what's actually on the page.
That's the whole case for getting a review before you pay for fabrication.
The obvious mistakes are easy to catch on your own. The expensive ones are the mistakes you don't know to look for.
Inside the Hardware Academy, experienced engineers review schematics, PCB layouts, and enclosures like this before you spend money on fabrication.
Talk soon,
John Teel
Predictable Designs
|
|
|
|
|
|
|
