Skip to content
Notes from the bench

Insights

Written by the engineers who did the work, not by a marketing team describing it.

Nothing published yet. Three are being written by the engineers who ran the projects, and they are listed below.

If one of them is the problem in front of you right now, do not wait for the article — ask the question directly. You will get an answer from the person who would have written it, and sooner.

In preparation

The first three

Not a content calendar. Three projects where somebody learned something worth writing down, and is writing it down.

  1. High-speed PCBIn preparation

    Propagation-delay skew across an 18-layer Gen4 stack-up

    Length matching is the part everyone knows about. The article is about what is left after the lengths agree — dielectric variation between layers, via stub resonance, and why a pair routed on two different layers does not arrive together even when the ruler says it should.

    The project behind it
  2. Low-power designIn preparation

    Where the current actually goes in a sub-5 µA node

    A datasheet sleep figure is the floor, not the budget. The article walks a real measurement session: what a current probe shows that a spreadsheet does not, which leakage paths dominate at single-digit microamps, and how much of the budget the radio was never responsible for.

    The project behind it
  3. Motor controlIn preparation

    Why open-loop ramping fails for elevator traction motors

    Open loop is cheap, well understood, and fine until the load is a lift car with people in it. The article covers what changes when load torque is not known in advance, why the failure appears at the worst moment rather than during commissioning, and what sensorless control has to establish before it can take over.

    The project behind it
How these get written

Why there are three and not thirty

Written by the engineer who did the work

Not by a marketing writer interviewing them. The difference shows up in the second paragraph, in whether the piece can say what was tried and abandoned.

No gate, no form, no PDF

The article is the page. Anything worth writing down is worth reading without giving us an email address first.

Published when it is true, not when it is due

There is no content calendar. An article appears when someone has finished a project they learned something from — which is why there are three planned and not thirty.

Ask the question now.

The engineer who would write the article is the one who would answer you. You will get a real answer rather than a discovery call — and you will get it this week rather than whenever the piece is finished.