E

Version control for hardware

Every part choice.
Every reason. Tracked.

EngTrace logs the reasoning behind every part, schematic change, and board revision. Six months later, nobody has to guess why.

Sensor board rev 412 decisions

3.3V LDO instead of a buck converter

Quieter analog rail for the ADC front-end.

rev 4

USB-C receptacle, 16-pin footprint

Confirm mechanical fit before the next spin.

rev 4

STM32H743 over the H750 for this rev

Extra RAM headroom for the FFT buffer.

rev 3

Where hardware knowledge goes to die

  • A Slack thread nobody can find six months later
  • A spreadsheet that only makes sense to the person who wrote it
  • The one engineer who remembers why, and they just left

Capture the why, not just the what

Every decision gets a reason, a status, and a timestamp. Mark it confirmed, tentative, or revisit, and change your mind later without losing the history.

3.3V LDO instead of a buck converter

Quieter analog rail for the ADC front-end. Efficiency loss is acceptable at this current draw. Revisit if the thermal budget on rev 5 gets tight.

Built for the whole board, not just the log

Parts and files all live next to the decisions that explain them.

One source of truth for every part

Quantities, prices, and status live next to the decision that put them there.

PartQtyStatusPrice
STM32H743VIT61Confirmed$12.40
TPS7A4901 LDO2Confirmed$1.85
USB-C receptacle, 16-pin1Revisit$0.62

Schema, code, and datasheets, together

View the files your decisions actually reference, without leaving the project.

Schema
Code
Datasheets
// sensor-board.c
#define ADC_VREF_MV 3300
#define FFT_SIZE 1024

void adc_init(void) {
// STM32H7, extra RAM for the FFT buffer
HAL_ADC_Start_DMA(&hadc1, buf, FFT_SIZE);
}

Trace decisions like a schematic

The schema view lays out every decision as a connected node, from part choice to board revision.

Built-in engineering calculators

Voltage dividers, trace width, RC filters. Reach for them without leaving the project.

One board, five decisions, no memory required

This is a real log: the same sensor board from a first pick through a flagged problem to the fix that shipped.

rev 1confirmed

Chose the STM32H750 for the sensor board

Plenty of headroom for the FFT buffer at this sample rate.

rev 2confirmed

Added a buck converter for the 3.3V rail

Efficiency mattered more than noise floor at this stage.

rev 3revisit

Buck converter is coupling noise into the ADC

Flagged during bring-up. Needs a quieter rail before the next spin.

rev 4confirmed

Swapped to a linear LDO

Traded a few points of efficiency for a clean analog rail. See rationale.

rev 4tentative

USB-C receptacle, 16-pin footprint

Looks right on paper. Confirm the mechanical fit before ordering boards.

Stop losing the reasoning.

Free to start. No credit card, no seat limits while you're finding out if this fits how you work.