You spend weeks designing a PCB and send it out for prototyping. The boards come back — something's off. So you revise, send it out again, and wait another week.
You tell yourself: next time, first-pass success.
Then next time comes, and you're revising again.
This isn't just you. Industry data shows that even for experienced engineers, the first-pass success rate for PCBs is only about 65%. Forcing "first-pass success" is itself a flawed goal. The real question isn't "how to get it right the first time" — it's "why is revision the norm, and how do we reduce it?"
The Real Reasons Boards Keep Coming Back
1. Vias That Are "Too Thin": The 0.2mm Domino Effect
A real-world case: a 20-board small-batch order used 0.2mm vias, adding RMB 232 to the cost — RMB 119 for the via size itself and another RMB 111 for associated four-wire low-resistance testing. Switching to 0.3mm would have eliminated both charges.
0.2mm vias aren't just more expensive — they're more failure-prone. Smaller drill bits wear faster and break more often. The plating solution struggles to penetrate the hole wall uniformly. Under high temperature or vibration, the copper walls are more likely to crack.
The revision cost: Either the factory charges extra for mandatory testing, or you wait for the boards to come back with cracked via walls and re-spin the entire batch.
2. Design Reviews That Aren't "Thorough Enough": One Silkscreen Mistake Can Kill a Board
What's the most common PCB design error? Industry statistics show:
The most frustrating part? Most of these don't show up in schematic simulation — they're hidden in the layout, only to be discovered when the boards arrive.
Example: an 0805 resistor footprint drawn at 0402 size. Or a through-hole capacitor with a pad diameter 0.2mm smaller than the lead. Components either don't fit, or the solder joints crack at the slightest stress.
And right-angle traces? To a high-frequency signal, they're "signal reflectors" — the signal hits the corner and bounces back, colliding with the next signal and turning the waveform into noise.
The revision cost: Re-spin, re-assemble, re-wait. Days to weeks.
3. Process Problems Run Deeper Than Technical Ones
There's a more subtle culprit: some teams' review processes are themselves designed to generate revisions.
One hardware engineer revealed that their manager insisted on "first-pass success" for all PCB designs — but the manager's review method was to propose numerous changes regardless of design quality, including non-essential adjustments. Each revision cycle took 3-7 days.
The result? Despite spending significant time on design verification, the first-pass success rate remained at zero. Common failures included impedance mismatches (42%), thermal design flaws (28%), EMC issues (19%), and others.
The revision cost: Not a technical problem — it's a process problem. The review itself has become the source of revisions.
How to Reduce Revisions
1. Introduce DFM Checks Instead of "Visual Inspection"
Many teams still rely on "eyeballing" designs. But the human eye misses over 15% of defects. The better approach: systematic DFM checks — line widths, clearances, via sizes, silkscreen placement — catching issues before they go to fab.
Industry data shows that early simulation catches 78% of electrical issues.
2. Accept That "First-Pass Success" Is Unrealistic
The industry average first-pass success rate is about 65%. Instead of chasing an unrealistic "one shot," build in tolerance for iteration. Spending 10 extra minutes on DRC before ordering can save 10 days of rework.
3. Prevent Errors at the Source
Most PCB errors are avoidable from the start:
- Check component datasheets before drawing footprints — don't guess
- Use 45° or curved traces instead of right angles
- Keep silkscreen at least 0.5mm from pads — don't let text "steal" the solder joint's position
The Bottom Line
PCB revisions aren't proof of "bad design" — they're a normal part of hardware development. The real problem isn't "getting it right the first time," but "making each revision count instead of stepping on the same rake twice."
Instead of chasing an unrealistic "first-pass success," build revision costs into your product design timeline — and use DFM check tools to catch avoidable errors before they reach production.
AnyPCBA, founded in 2011, focuses on small-to-medium batch PCB manufacturing and PCBA assembly (5–5,000 pieces). If you're not sure whether your design has hidden manufacturing risks, send us your Gerber files — we'll find the "landmines" before you order.
👉 AnyPCBA website: https://www.anypcba.com/
Small-to-medium batch PCB & PCBA | 5–5,000 pieces | Prototype to Production
Top comments (0)