Insights · Reverse engineering
When to reverse engineer obsolete hardware — and when to redesign
BLINKENLIGHT · ~6 min read
A board that has run a process line for fifteen years finally dies. The supplier is dissolved. The design files never existed in any form you can find. The microcontroller went end-of-life two revisions ago. The question is rarely "can this be fixed". It is "which route is least risky for the next ten years". There are three routes, and choosing well saves far more than the engineering costs.
Route 1 — Recover the documentation
Reverse engineer the existing board into a verified schematic, bill of materials and, where needed, layout. You keep the board's behaviour identical. You turn an undocumented board into a maintainable asset. This is the right first move when the design is still sound. It fits when the parts are mostly available. The real problem is lost knowledge. Recovery restores that knowledge. It is also the cheapest route, and it sets up the other two.
Route 2 — Build a compatible replacement
Recover the design, then re-spin it to remove the obsolete parts. Keep the same form, fit and function. The connectors, mounting, timing and interfaces stay identical, so the surrounding system sees no change. This is usually best value when only a handful of components are the problem. The wider system stays untouched. The work is bounded and the validation target is clear. It must behave indistinguishably from the original.
Route 3 — Redesign
Treat the obsolescence as the trigger for a deliberate new design. Redesign is right only when the original genuinely limits you. The limit may be performance, cost, certification or supply. You also need to absorb a full validation cycle. Plan it as a redesign from the start, and the schedule holds.
A quick decision rule
- Documentation lost, hardware fine, parts available → recover.
- A few obsolete parts, system must not change → compatible replacement.
- The original limits the product and you can re-qualify → redesign.
Two things decide the cost of all three routes. The first is how much of the original is RF or analogue. Those sections demand exact reproduction. The second is whether the result must support a regulatory or market-entry file. Establish both before anyone touches a soldering iron.
Where this connects
Most obsolescence work we do starts as Route 1 and resolves into Route 2. A short scoping conversation settles the decision. See product & PCB reverse engineering for how we run it, including IP boundaries and the documentation deliverables.
Book a free 30-minute consultation →
Published May 2026 · All insights