Industrial automation and HMI
Field wiring brings surges and ground-potential differences into your enclosure, and the product has to still be buildable in ten years. Both constrain the design more than the function does.
Evidence
What we can substantiate
Everything below is work we have actually done, described at a level that respects the client’s confidentiality. If you need detail, ask under NDA.
- Galvanic isolation on every field-facing interface Two panels a hundred meters apart do not share a ground. The potential difference drives current through any non-isolated interface and destroys transceivers — usually intermittently and usually blamed on noise. Isolated RS-485 and CAN with their own supply domain, and the isolation barrier respected in the layout rather than only in the schematic.
- Surge coordination, not just a TVS on the connector IEC 61000-4-5 surge testing fails when a fast clamp is placed downstream of a slower one, or when the clamp voltage is above what the transceiver survives. We coordinate the chain — gas tube or MOV, series impedance, then the semiconductor clamp — and check the let-through against the part it protects.
- Wide-temperature derating at both ends Cold failures are as common as hot ones and far less anticipated: electrolytic ESR rises steeply below zero, crystals drift out of a PLL's capture range, LCDs slow. We derate to the specified minimum as deliberately as to the maximum.
- Longevity engineered rather than hoped for A ten-year product needs parts with a stated longevity program, a second source identified at design time for anything critical, and no dependence on a package or a die revision that is already mature. Lifecycle status is checked at selection, not at the first EOL notice.
- Industrial Ethernet and TSN where determinism matters EtherCAT, PROFINET and TSN place real constraints on PHY selection, clock distribution and PCB routing — jitter budgets are tight and a marginal layout shows as intermittent dropped cycles, not as a clean failure.
- Conformal coating specified with rework in mind Condensing environments need coating, and the choice of chemistry decides whether the board can ever be repaired. Decided before layout so keep-out areas, connector masking and test points are placed accordingly.
Common questions
Straight answers
Why do RS-485 transceivers keep failing in the field?
Usually ground-potential difference rather than surge. Two cabinets on different distribution boards can sit volts apart at mains frequency, and more during a fault. A non-isolated transceiver carries that current through its ground pin until it fails — often weeks in, and often blamed on lightning. Galvanic isolation with a separate supply domain fixes the cause; a bigger TVS only delays it.
How do you handle a ten-year availability commitment?
By selecting for it. Parts from vendors with a published longevity program, second sources identified during design rather than during a shortage, and an architecture that does not depend on one specific package. We also record why each critical part was chosen, so when it does go end-of-life the person handling it knows what the replacement has to satisfy. Ten years is long enough that some obsolescence is certain — the goal is that it is survivable rather than avoided.
Do you design to IEC 61508?
We design hardware that fits into a functional-safety program and we can carry the hardware-side analysis — failure modes, diagnostic coverage, architectural constraints. Certification of the overall safety function involves an assessor and a lifecycle that sits with you or a specialist partner. If your product needs an SIL rating, tell us at the start, because it changes the architecture rather than the documentation.
Got a board to design — or one that won’t boot?
You talk to the engineer who would do the work. Reply within one business day, and we’ll sign your NDA before you go into detail.