What Does a 5000 W TVS Rating Actually Mean? Peak Pulse Power, Pulse-Width Derating and Average Power
One-line takeaway: the 5000 W on a TVS datasheet is peak pulse power (Pppm) — clamping voltage × peak current under a 10/1000 μs standard waveform with a single non-repetitive pulse. It is not an 8/20 μs figure, and it is certainly not a continuous steady-state rating. Select in three steps: derate by pulse width from the actual surge waveform → estimate average power by duty cycle for repetitive pulses (typically only a few watts) → derate again for high ambient temperature. Treating the nominal power as what the part "can handle" is the most common waveform-convention error in TVS application. AMSEMI 5.0SMDJ datasheets (8060014 / 8060055) use only the 10/1000 μs waveform throughout, with no 8/20 μs data.
flowchart TD
A["Define the surge event:<br/>waveform, pulse width, repetition rate"] --> B{"Single occasional pulse<br/>or repetitive pulses?"}
B -- "Single pulse" --> C["Derate by actual pulse width<br/>10/1000 us nominal as baseline<br/>shorter pulse allows higher peak"]
C --> D["Check Ipp and Vc<br/>Vc x Ipp must stay under derated power"]
B -- "Repetitive" --> E["Estimate average power by duty cycle<br/>P_avg order = P_peak x tau/T<br/>check junction temp accumulation"]
E --> F["Derate again for high ambient<br/>per datasheet temperature curve"]
D --> F
F --> G["Return to the three-stage voltage window<br/>Vc must still fit the downstream budget"]Two Questions Answered First
Q: What does a 5000 W TVS rating mean? Can it really handle 5000 W continuously?
No. Peak pulse power Pppm is defined as the product of clamping voltage and peak pulse current during one surge of the specified waveform: Pppm = Vc × Ipp. For the 85 V bidirectional grade of a 5000 W series at 10/1000 μs: Vc = 137 V, Ipp ≈ 36.5 A, and 137 V × 36.5 A ≈ 5000 W. The nominal number is that product under that one test condition. It answers "how much energy can one surge absorb" and has nothing to do with steady-state watts: converted to average power, the same part's continuous capability is typically only a few watts.
Q: I test to IEC 61000-4-5 at 8/20 μs. Can I convert 5000 W into an Ipp?
Not directly. 10/1000 μs and 8/20 μs distribute energy very differently: the former rises in 10 μs and decays to half value in 1000 μs — long pulse, high single-event energy; the latter rises in 8 μs and decays to half in 20 μs — the short-circuit current waveform of the lightning-protection combination wave generator. The same device tolerates far higher peak current under the short waveform. On Littelfuse published data for the same-package part 5.0SMDJ85CA, Ipp is 36.5 A at 10x1000 μs while the separately listed IPP 8x20 μs is 273.8 A (Littelfuse's own device data), about 7.5× apart. AMSEMI 5.0SMDJ datasheets (8060014 / 8060055) contain only the 10/1000 μs convention and no 8/20 μs data — comparing an 8/20 μs test result directly against AMSEMI datasheet numbers is a convention mismatch.
The Three Conditions Behind the Nominal Number
The large Pppm on page one holds only when all three are true. Miss one and the number cannot be used directly:
| Condition | Meaning | Common misreading |
|---|---|---|
| Waveform convention | 10/1000 μs standard waveform (common for telecom power and BMS TVS) | Comparing an 8/20 μs test result directly; convention mismatch |
| Single, non-repetitive | One surge, with heat from any previous pulse fully dissipated | Using single-pulse power as repetitive capability; junction temperature runs away |
| Reference temperature | Usually 25 °C starting junction temperature (confirm per manufacturer) | No derating at high ambient; usable power falls below the nominal figure |
Two Common Waveforms: Do Not Mix the Numbers
| Waveform | Rise / half-value time | Typical use |
|---|---|---|
| 10/1000 μs | 10 μs / 1000 μs; long pulse, high single-event energy | Conventional TVS datasheet waveform for telecom power and BMS |
| 8/20 μs | 8 μs / 20 μs; short pulse, high peak current | IEC 61000-4-5 combination wave short-circuit current; lightning protection |
Two hard rules applied consistently across this site:
- Vc and Ipp for the AMSEMI 5.0SMDJ series are entirely 10/1000 μs (datasheet 8060014 standard / 8060055 foldback; no 8/20 μs data anywhere).
- Littelfuse datasheets list IPP 8x20 μs separately (e.g. 5.0SMDJ85CA: 10x1000 μs 36.5 A / 8x20 μs 273.8 A). That is Littelfuse's own device data and must be explicitly attributed to Littelfuse, never written as "per the AMSEMI datasheet".
Clamping voltages from the two conventions cannot be compared by subtraction. For cross-convention comparison use only relative measures such as percentage reduction, and state the test condition on every row. See Surge Protection for Energy Storage and BMS for how waveform conventions map onto system protection levels.
Pulse-Width Derating: Longer Pulse, Lower Allowed Peak
The nominal power corresponds to exactly one test condition, 10/1000 μs. For any other pulse width, convert using the datasheet pulse-width derating curve:
- Pulse shorter than 1000 μs (tens-of-microseconds spikes): the same part tolerates more than the nominal peak power; short-pulse peak current capability rises substantially (see the roughly 7.5× gap between Littelfuse 8x20 μs and 10x1000 μs above).
- Pulse longer than 1000 μs (multi-millisecond slow surges, inductive load switch-off): allowed peak power falls below nominal and must be derated by the curve; available power shrinks sharply at long pulse widths.
- How to read it: pulse width on the horizontal axis, allowed peak power on the vertical; find your actual pulse width, then stack temperature derating on top.
For long-pulse inductive switch-off energy accounting, see How to Suppress Relay Coil Turn-Off Spikes .
Repetitive Pulses: Average Power Is Usually Only a Few Watts
Repetitive surges cannot use the single-pulse convention. Estimate average power from the duty cycle:
P_avg ≈ P_peak × (pulse width τ ÷ repetition period T)
Example: 5000 W single pulse, 1 ms width, once per second (T = 1 s) → about 5000 W × 0.001 = 5 W. That is the intuitive reason a TVS's "continuous capability" is only a few watts. Then check two more things:
- Junction temperature accumulation: convert average power to steady-state temperature rise through thermal resistance, add ambient, and confirm it stays below maximum junction temperature (thermal parameters per datasheet).
- Datasheet blind spot: most TVS datasheets give only the 10/1000 μs single-pulse rating with no repetitive rating. For dense repetitive surge applications — repeated reverse-polarity hot plugging, frequently switching relays — send the pulse train conditions (width, peak, frequency, duty cycle) to the manufacturer for confirmation, and measure temperature rise if necessary.
Temperature Derating: Pulse Width First, Then Temperature
Nominal power is usually referenced to a 25 °C starting junction temperature (some manufacturers start at 20 °C or other points; confirm per part number). How to apply temperature derating:
- Determine the starting junction temperature when the surge arrives, accounting for heat not yet dissipated from a previous pulse;
- Read the usable power ratio at that temperature from the datasheet temperature derating curve — higher temperature means lower usable peak;
- Pulse-width and temperature derating stack: convert for pulse width first, then for temperature, citing the basis at each step.
Selection Checklist
| Check | What to look at | Basis |
|---|---|---|
| Waveform convention matches | Datasheet convention vs. actual test waveform | Test conditions on datasheet page one; AMSEMI is 10/1000 μs |
| Single-pulse energy | Derated power at actual pulse width ≥ surge power | Pulse-width derating curve |
| Repetitive pulses | P_avg ≈ P_peak × τ/T, then check junction temperature | Thermal resistance + manufacturer confirmation |
| High-temperature derating | Ratio at the starting junction temperature | Temperature derating curve |
| Clamping window | Vc must still fit the downstream withstand budget | Three-stage voltage window |
That last row is where power conventions most often mislead: passing the power check does not mean protection is achieved — clamping voltage still has to be checked against the three-stage voltage window (standoff, breakdown, clamping). See Why a BMS Board Still Blows MOSFETs With a TVS Installed: The Voltage Budget for 23/24-Cell Packs .
Common Pitfalls
| Pitfall | Symptom | Correct approach |
|---|---|---|
| Treating nominal power as steady-state capability | "How did a 5000 W part cook its junction at a few tens of watts?" | Estimate average power by duty cycle; typically only a few watts |
| Comparing numbers across conventions | An 8/20 μs test result never matches the AMSEMI datasheet | Unify the convention first; AMSEMI is 10/1000 μs with no 8/20 μs data |
| Applying single-pulse rating to repetitive surges | Junction temperature accumulates, device degrades early | Estimate average power plus thermal resistance; send pulse train to the manufacturer |
| No derating at high temperature | Failure rate rises sharply in summer sites or sealed enclosures | Derate by the temperature curve, stacked with pulse-width derating |
Related Reading
- BMS selection overview: How to Select TVS for BMS Protection Boards
- Troubleshooting power-convention mistakes: Why BMS Protection Boards Still Fail With a TVS Installed
- Three-stage voltage window: The Voltage Budget for 23/24-Cell Protection Boards
- System-level surge grading: Surge Protection for Energy Storage and BMS
- Full series: AMSEMI Foldback TVS: Full Series and Cross-Reference
Frequently Asked Questions (FAQ)
Q1: What does a 5000 W TVS rating mean, and can it handle 5000 W continuously?
No. The Pppm on page one is defined under three conditions: 10/1000 μs waveform, single non-repetitive pulse, and the product of clamping voltage and peak pulse current. For the 85 V grade at 10/1000 μs, 137 V × 36.5 A ≈ 5000 W. It is not an 8/20 μs figure and not a steady-state capability; continuous capability must be computed from average power and thermal resistance and is typically a few watts.
Q2: What is the difference between 10/1000 μs and 8/20 μs, and which does AMSEMI use?
10/1000 μs is a long pulse (10 μs rise, 1000 μs to half value), the conventional TVS waveform for telecom power and BMS. 8/20 μs is the IEC 61000-4-5 combination wave short-circuit current waveform used in lightning protection. The same part allows very different peak currents: Littelfuse lists 36.5 A at 10x1000 μs and 273.8 A at 8x20 μs for 5.0SMDJ85CA, about 7.5× apart, and that 8x20 figure is Littelfuse's own data. AMSEMI datasheets 8060014 and 8060055 use only 10/1000 μs with no 8/20 μs data.
Q3: How do I handle repetitive pulses and estimate average power?
Estimate average power by duty cycle: P_avg ≈ P_peak × τ / T. A 5000 W pulse of 1 ms repeating once per second gives about 5 W average. Then verify that the junction temperature rise stays inside the safe area using datasheet thermal resistance, and note that most datasheets carry no repetitive rating — for dense repetitive surge applications, send the pulse train conditions to the manufacturer.
Q4: Does a TVS need derating at high ambient temperature?
Yes. Nominal power is usually referenced to a 25 °C starting junction temperature. Determine the starting junction temperature when the surge arrives (including residual heat from prior pulses), read the usable power ratio from the temperature derating curve, and check surge energy against nominal × ratio. Pulse-width and temperature derating stack: pulse width first, then temperature.
Sources
- AMSEMI 5.0SMDJ series datasheets: doc. 8060014 (standard) and doc. 8060055 (-N foldback), 10/1000 μs only
- Littelfuse published parameter table for 5.0SMDJ85CA (10x1000 μs 36.5 A; separately listed 8x20 μs 273.8 A)
- IEC 61000-4-5 combination wave definitions
Contact Us
Need to reconcile a TVS power convention against your actual surge waveform and pulse train, or samples of the AMSEMI 5.0SMDJ series? Contact us — send the test waveform, pulse width and repetition rate and we will balance the numbers against the original datasheet convention.
Shenzhen Intek Technology Co., Ltd — electronic components distributor and system solutions provider Tel / WeChat: 136-3264-8484 Address: 4F, Building 2, Jingwei Center, No. 309 Pingan Avenue, Pinghu Street, Longgang District, Shenzhen, China Web: www.intek.vip
Disclaimer: This article is technical education and industry discussion for reference only and does not constitute selection, procurement or other commercial advice. Peak pulse power, clamping voltage, peak current and derating curve information is taken from published manufacturer material (AMSEMI datasheets 8060014 / 8060055, Littelfuse published parameter tables) and public standards. Average power and derating examples are order-of-magnitude estimates; actual values vary significantly by part number, waveform and operating condition, and designs must be verified against the official datasheet for the selected part and measured on the real board.