A field team faces an unresolved physical question: Why can a pump reset a board before firmware gets a vote? They must answer it before changing relay current on the real device. Predict the direction first.
See the relationship before changing it
The figure reads from left to right. The blue card is relay current. The middle card applies this page's relationship. The green card is pump sag. Walk the arrows once: set the input, apply the rule, then read the result with its unit.
The retained audit below checks several chapter fixtures. This added model holds every other chapter fixture fixed, so the numeric fixture does not switch without explanation.
Derive the baseline in four named moves
- 1
Name the input. The chapter baseline for relay current is 1.5.
- 2
Name the relationship. Icoil=5/71=70.4 mA ΔVpump=1.50x2.00=3.00 V Vrail=5.00-3.00=2.00 V Regulation threshold=3.30+0.30=3.60 V VL=0.3x0.0704/1 us=21,127 V
- 3
Substitute the chapter fixture. Set relay current to 1.5. The page ledger gives pump sag as 3.00 V.
- 4
Read the result. Keep V beside the value. Use it only inside the technical boundary on this page.
Predict, then change relay current
Try Predict the direction of pump sag. Move one control, calculate, then check your prediction.
Observe The board resets when the electrical state leaves its allowed region; firmware is downstream of that event. Reset the control to 1.5 and compare pump sag.
Explain Only relay current moves here. The other chapter fixtures remain fixed.
Check yourself
What should you do before trusting a moved-control result?
What does this small model leave out?
1. Start with current and resistance
Current is charge flowing each second. Resistance makes that flow spend voltage. A source wire, connector, breadboard rail, and ground return all contribute resistance even when no resistor symbol appears in the diagram.
2. Name every algebra move
Find the coil currentI=V/R.
Find the shared-path dropΔV=IinrushRsource.
Subtract from the sourceVrail=Vsupply−ΔV.
Test regulationCompare Vrail with Vout+Vdropout.
Estimate unclamped kickVL≈LI/Δt.
3. Reproduce the chapter case
ΔVpump=1.50×2.00=3.00 V
Vrail=5.00−3.00=2.00 V
Regulation threshold=3.30+0.30=3.60 V
VL=0.3×0.0704/1 µs=21,127 V
The 2.00 V rail is 1.60 V below the regulator-input threshold and 0.43 V below the named 2.43 V brownout level. The huge flyback number describes an intentionally unclamped teaching estimate, not the voltage across a fitted diode.
4. Try the pump inrush
TryMove the pump inrush while the supply and shared path stay fixed.
ObserveEvery extra 0.1 A removes another 0.20 V because the 2.00 Ω path has not changed.
ExplainThe board resets when the electrical state leaves its allowed region; firmware is downstream of that event.
This is a lumped steady-path and unclamped-coil ledger, not a transient circuit simulator.
- Inrush
- Real pump current changes with load, driver, wiring, and time
- Regulator
- Dropout and brownout thresholds depend on the exact parts and configuration
- Flyback
- A diode, TVS, transistor avalanche, capacitance, and parasitics limit the real voltage
Capture the rail and switch node with suitable probes, then verify the fitted protection network.
5. Test the fault in the right order
Measure the 5 V input, 3.3 V rail, pump current, and reset line on one time base. Then separate the pump supply or shorten the return path and repeat the same start event.
6. Record the evidence state
Store the supply, cable, connector, pump load, relay part, probe points, time scale, minimum rail, protection parts, firmware hash, and the single change that removed or preserved the reset.
7. Check yourself
Why does 1.50 A through 2.00 Ω remove 3.00 V?
Why is 2.00 V not enough for a 3.3 V regulator?
Does the 21,127 V estimate predict the clamped switch voltage?
The arithmetic reproduces the chapter's catalog-typical pump, relay, path, regulator, and radio example.
- 2.00 V
- One shared-path estimate, not a measured board rail
- 21,127 V
- An unclamped idealization, not a safe probe expectation
- 1.96 V
- A stacked teaching result, not a complete power-distribution model
Correct, not complete: this ledger does not diagnose a particular reset without a captured waveform.
Voltage Vera guides