A field team faces an unresolved physical question: Why does 400 kHz I2C need a stronger pull-up? They must answer it before changing i2c clock 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 i2c clock. The middle card applies this page's relationship. The green card is clock period. 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 i2c clock is 400.
- 2
Name the relationship. T400k=1/400000=2.50 us tr,budget=0.10(2.50 us)=250 ns Rmax=250 ns/(0.847x100 pF)=2.95 kohm Isink=3.3/2952=1.12 mA T100k=10.0 us; Rmax=11.8 kohm
- 3
Substitute the chapter fixture. Set i2c clock to 400. The page ledger gives clock period as 2.50 us.
- 4
Read the result. Keep us beside the value. Use it only inside the technical boundary on this page.
Predict, then change i2c clock
Try Predict the direction of clock period. Move one control, calculate, then check your prediction.
Observe The pull-up cannot be selected from rise time alone: it must be low enough to charge C, yet high enough that every device can sink the resulting current. Reset the control to 400 and compare clock period.
Explain Only i2c clock 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 the physical story
An I2C device pulls a line LOW, but no device actively drives it HIGH. Releasing the line lets the pull-up resistor charge all bus capacitance. A faster clock leaves less time for that charge.
2. Name every algebra move
Invert clock frequencyT=1/f gives one period.
Reserve edge timeMultiply T by the stated 10% illustration.
Use threshold riseFor 30% to 70%, tr≈0.847RC.
Isolate the resistorR=tr/(0.847C).
Check LOW currentWhen asserted, I≈VDD/R.
3. Reproduce the chapter case
tr,budget=0.10(2.50 µs)=250 ns
Rmax=250 ns/(0.847×100 pF)=2.95 kΩ
Isink=3.3/2952=1.12 mA
T100k=10.0 µs; Rmax=11.8 kΩ
Quadrupling the clock divides the period, rise budget, and allowed resistance by four while multiplying the pull-up current by four.
4. Try one real input
TryMove clock speed and predict the largest pull-up allowed by the same illustrative budget.
ObserveMore clock speed shrinks allowable R and raises sink current. The UART byte time does not change because it belongs to a different push-pull timing contract.
ExplainThe pull-up cannot be selected from rise time alone: it must be low enough to charge C, yet high enough that every device can sink the resulting current.
The 10% budget and 100 pF bus are explicit illustrations, not universal I2C limits.
- Timing
- Use the selected I2C mode's specified rise time and each device's timing limits.
- Capacitance
- Trace, cable, connector, probes, and every device pin contribute to the real C.
- Current
- VOL, sink rating, resistor tolerance, voltage, and leakage bound the strong-pull-up side.
Correct, not complete: this ledger does not certify 400 kHz operation or choose a production pull-up.
5. Use the result in the design
Measure total capacitance or the actual edge, apply the mode's rise-time limit, then check VOL and sink current across voltage and temperature.
6. Record the evidence state
Record clock mode, VDD, pull-up value and tolerance, device count, estimated and measured capacitance, measured rise time, VOL, sink current, harness, and probe loading.
7. Check yourself
Why does 400 kHz have one quarter of the 100 kHz period?
Why can a pull-up be too strong?
Does 2.95 kΩ prove this bus works?
The arithmetic reproduces the chapter's 100/400 kHz inverse scaling with an explicitly illustrative 100 pF, 10% edge budget.
- Timing
- Use the selected I2C mode's specified rise time and each device's timing limits.
- Capacitance
- Trace, cable, connector, probes, and every device pin contribute to the real C.
- Current
- VOL, sink rating, resistor tolerance, voltage, and leakage bound the strong-pull-up side.
Correct, not complete: this ledger does not certify 400 kHz operation or choose a production pull-up.
Eddie guides