Tantalum Capacitor Derating Calculator
Tantalum Capacitor Derating Calculator
Tantalums fail short and a shorted one in a low-impedance circuit can ignite, and every published tantalum design rule follows from that. Voltage derating, the series resistance rule in ohms per applied volt, the turn-on surge current, the category voltage above 85 °C, and what changes when the cathode is a conductive polymer instead of manganese dioxide.
Voltage derating, the ohms-per-volt rule and the turn-on surge
a 10 µF 16 V MnO₂ tantalum on a 5 V rail at 105 °C, with 50% voltage derating, a category temperature of 85 °C and two thirds of the rating left at 125 °C
Four rules, and they are not the same rule
combined limit = Vcat(T) × D
Rrequired = (Ω per volt) × Vworking
Isurge = 1.1 VR ÷ (Rcircuit + ESR) τ = (Rcircuit + ESR)·C
- D
- the voltage derating your programme requires. 50% is the widely applied figure for solid tantalum in a low-impedance circuit; KYOCERA AVX’s own paper argues the number belongs to the application rather than to the part
- Ω per volt
- the series resistance rule. It bounds fault current, so it bounds the energy a defect site can see. Historically 3, then 1, then 0.1 ohms per applied volt or 1 ohm, whichever is greater
- 1.1
- the factor in KYOCERA AVX’s published surge screening expression — the screen is applied above rated voltage
- ESR
- in series with the circuit resistance, so on a stiff rail it is the capacitor’s own ESR that limits the inrush. A polymer part’s much lower ESR removes part of that limit
Worked example
a 10 µF 16 V MnO₂ tantalum on a 5 V rail at 105 °C, with 50% voltage derating, a category temperature of 85 °C and two thirds of the rating left at 125 °C
At 105 °C the part is 20 °C into a 40 °C span, so the maximum working voltage has fallen to 83.34% of the rating: 13.33 V rather than 16 V
The 50% derating applies to that, giving a combined limit of 6.667 V. A 5 V rail is 75.0% of it — where derating the bare rating would have given 8 V and made the same rail look like 62.5%
The series resistance rule is the other requirement and it is independent of that one. At 1.0 Ω per applied volt a 5 V rail asks for 5 Ω; the 500 mΩ in the turn-on path is 0.10× that, so the largest working voltage this circuit impedance permits under the rule is 500 mV
Turn-on into a discharged part sees 500 mΩ plus the capacitor's own 900 mΩ of ESR, so the inrush peaks at 5 V ÷ 1.4 Ω = 3.571 A and decays with a 14 µs time constant. KYOCERA AVX's screening expression at rated voltage, 1.1 × 16 ÷ 1.40, gives 12.57 A
And the ripple: 0.15 A rms through 900 mΩ is 20.25 mW, which is 18.4% of the 110 mW the part allows — and every watt of it raises the temperature that set the category voltage in the first line
The four tantalum rules and what each is actually about
| Rule | What it protects against | Where it comes from |
|---|---|---|
| Voltage derating, commonly 50% for MnO₂ in a low-impedance circuit | the field across the dielectric during a surge, and whether a scintillation at a flaw self-heals or runs away. Not lifetime | KYOCERA AVX, Voltage Derating Rules for Solid Tantalum and Niobium Capacitors — which argues the rule belongs to the application: 50% where a battery feeds the part directly, as little as 20% where the source impedance is controlled |
| Series resistance, in ohms per applied volt | fault current, and therefore the energy a defect site can dissipate | Vishay characterise commercial MnO₂ parts with 1 Ω per applied volt; NASA NEPP records the rule as 3 Ω/V in the 1960s, 1 Ω/V by the 1980s, 0.1 Ω/V or 1 Ω whichever is greater in the 1990s |
| Surge current at turn-on | the current step into a discharged capacitor, which is bounded only by the circuit resistance and the ESR | KYOCERA AVX give the screening peak as 1.1·V_R ÷ (R + ESR); the surge screen applies one or more current surges and watches the leakage decay afterwards |
| Category voltage above the category temperature | the dielectric at temperature, where leakage rises and the margin for self-healing falls | Vishay document 40246: rated voltage from −55 °C to +85 °C, then falling linearly to two thirds of it at the 125 °C maximum working temperature |
MnO₂ against conductive polymer
| Manganese dioxide | Conductive polymer | |
|---|---|---|
| Cathode | MnO₂ — an oxidiser, carrying its own oxygen | a conductive polymer, with much less oxygen in the cathode |
| Failure mode | short, and in a circuit that can supply current an ignition risk | Vishay state that polymer parts do not have an ignition failure mode |
| ESR | ohms to tenths of an ohm for a small part | tens of milliohms — KEMET quote 25 to 150 mΩ at 100 kHz for one polymer series |
| Effect on inrush | the part’s own ESR limits the turn-on current and helps | that help is largely gone; the circuit resistance has to do the work |
| Voltage derating | 50% is the widely applied figure, and the application decides whether it is enough | manufacturers generally recommend less; take the figure from your part’s data sheet rather than from a rule of thumb |
| Temperature | rated voltage to the category temperature, then a linear fall | the same shape, and families rated above 105 °C carry their own derating above it — KEMET’s T599 note gives two bands, 105 to 125 °C and 125 to 150 °C |
Why a tantalum has rules that no other capacitor has
A solid tantalum capacitor is a sintered tantalum slug with an amorphous Ta₂O₅ dielectric grown on it and a cathode laid over that. It is small, stable, and does not dry out. It also fails short rather than open, and in a manganese dioxide part the cathode is an oxidiser carrying its own oxygen — so a short in a circuit that can supply current is not merely a failure, it is an ignition risk. Every published tantalum design rule is a consequence of that one fact, and they are four separate rules rather than one.
Voltage derating is not about lifetime. This is the part most often got wrong. An electrolytic’s voltage derating buys life; a tantalum’s buys margin against a surge event that either happens or does not. What it is really protecting is the field across the dielectric at a flaw, and whether a scintillation there is self-healed by the cathode or instead runs away. KYOCERA AVX’s own paper on the subject is blunt about the consequence: the right derating belongs to the application, not to the part. Their worked examples run from 50% where a battery feeds the capacitor directly, to as little as 20% where the source impedance is controlled — and they say plainly that tantalums can be used at 80% of rated voltage if you will accept a worse failure rate and higher leakage.
The ohms-per-volt rule is the one nobody can explain. It is the most cited tantalum design rule and the least understood, and it is not a derating rule at all — it is a fault-current rule. A series resistance of so many ohms per applied volt bounds the current a short can draw, and therefore the energy a defect site can dissipate. Vishay characterise commercial MnO₂ parts at 85 °C at rated voltage with 1 Ω per applied volt in series. NASA’s NEPP programme has traced the rule’s history: 3 Ω per volt in the 1960s, 1 Ω by the 1980s, and by the 1990s 0.1 Ω per volt or 1 Ω, whichever is greater — it has loosened as the parts improved, and which version binds you is a programme decision rather than a physical constant. What has not changed is the shape of the argument: a correctly derated tantalum on a hard low-impedance rail with nothing in series is still the part that fails.
Surge current is where the rule bites in practice. Switch a supply onto a discharged capacitor and the current is bounded only by the total series resistance — the source, the switch, the track and the capacitor’s own ESR. KYOCERA AVX give the peak their screening test produces as 1.1·V_R ÷ (R + ESR), and the screen exists precisely because parts that survive it are the ones that ship. Note what that expression contains: on a stiff rail with a low-resistance switch, the ESR is the limit. Swap an MnO₂ part for a polymer one with a tenth the ESR and the inrush goes up by roughly that factor, in a circuit nobody changed.
Temperature, and how it compounds. Above the category temperature the maximum working voltage falls. Vishay’s derating note describes the standard MnO₂ case as rated voltage from −55 °C to +85 °C and then a linear fall to two thirds of it at 125 °C. Your programme’s derating factor applies to what that curve leaves, not to the nameplate, and the two compound — the same arithmetic the general component derating calculator does for every other part family. Ripple current closes the loop: it heats the part through its ESR, and a hotter part is further down the curve that set the voltage limit in the first place.
Where this page sits beside the other capacitor pages. A Class II ceramic’s problem is that the capacitance is not the marked value under bias — the MLCC DC bias calculator is about that, and it is a capacitance problem rather than a safety one. An aluminium electrolytic’s problem is wear-out driven by ripple heating — the ripple current and ESR calculator is about that, and it is a lifetime problem. A tantalum’s problem is none of those: the capacitance is stable and the part does not wear out in the same way. Its problem is the one event, and that is what this page computes. Use the figures from your part’s datasheet; typical values vary widely between manufacturers.
Frequently asked questions
Is the 50% rule for tantalums real?
It is real as practice and it is not a law. KYOCERA AVX’s own paper on voltage derating argues that the necessary derating depends on the application: 50% where a battery feeds the capacitor through almost nothing, as little as 20% where the source impedance is controlled, and they state that tantalums can be used at 80% of rated voltage at the cost of a worse failure rate and higher leakage. What your programme requires is in your programme’s document. Type it in above.
What is the ohms-per-volt rule and do I still have to follow it?
It asks for a stated series resistance per applied volt in the turn-on path, and it is a fault-current rule rather than a derating rule: it bounds the energy a defect site can dissipate. NASA’s NEPP survey records it as 3 Ω per volt in the 1960s, 1 Ω per volt by the 1980s and 0.1 Ω per volt or 1 Ω whichever is greater by the 1990s — it has loosened as the parts improved. Whether it binds you, and at what value, is a programme decision; the page computes the resistance the figure you enter demands and the largest working voltage the resistance you actually have would permit.
Do polymer tantalums need derating at all?
Yes, and less. Vishay state that polymer parts do not have an ignition failure mode because there is less oxygen in the cathode, and manufacturers generally recommend less voltage derating than for MnO₂. Take the figure from your part’s data sheet. Notice the other half of the swap: a polymer part’s ESR is a small fraction of an MnO₂ part’s, and the ESR was part of what limited the inrush current, so a like-for-like substitution increases the surge in a circuit nobody changed.
Why does the inrush current matter if it only lasts microseconds?
Because the failure mechanism is energy delivered into a microscopic defect in the dielectric, not average power. The peak is set by the total series resistance — source, switch, track and the capacitor’s own ESR — and on a stiff rail it can be tens of amperes. That is exactly why manufacturers screen parts with a surge test and why the ohms-per-volt rule exists.
Can I use a tantalum straight across a battery?
That is the worst case in KYOCERA AVX’s own examples: a source with almost no impedance, so nothing bounds the fault current. It is the configuration that produced the 50% rule. If you must, derate hard, put real resistance in series, and consider whether a polymer part or a different technology removes the question.
My part runs at 105 °C. Does the derating change?
Yes. Above the category temperature the maximum working voltage falls, and your derating factor applies to what that curve leaves rather than to the rating. The page shows both readings — the product of the two, and the smaller of the two, which is how some programmes write the requirement — and the chart shows the whole curve so you can see where your rail sits against it across the temperature range.
Related calculators
References
- KYOCERA AVX. Voltage Derating Rules for Solid Tantalum and Niobium Capacitors (CARTS Europe 2003 technical paper, still published on the manufacturer’s site). Argues that the prevalent minimum-50% guideline is a statement about the application rather than about the part, with worked examples from a battery feeding the capacitor directly to a controlled-impedance DC/DC output where as little as 20% suffices; states that tantalums can be used at 80% of rated voltage at the cost of MTBF and leakage; gives the surge screening peak as Ip = 1.1·V_R ÷ (R + ESR) and the screening circuit resistance as a maximum of 1 Ω for standard product and 0.7 Ω for low-ESR series.
- Vishay. Solid Tantalum Capacitors (With MnO₂ Electrolyte) Voltage Derating, document 40246, revision 07 September 2022. Recommends a working voltage between 0.5 and 0.6 of the rated voltage, and a higher derating ratio still for ratings of 35 V and above; describes the maximum working voltage as equal to the rated voltage from −55 °C to +85 °C and then falling linearly to two thirds of it at the 125 °C maximum working temperature; and characterises commercial-grade MnO₂ parts at 85 °C at rated voltage with a 1 Ω series resistor per applied volt.
- Teverovsky A. Derating of Surge Currents for Tantalum Capacitors, NASA Electronic Parts and Packaging (NEPP) programme, presented at ESTEC 2013. A US Government work. Records the history of the series resistance requirement — 3 Ω per volt of operating voltage in the 1960s, reduced to 1 Ω by the 1980s and to 0.1 Ω per volt or 1 Ω whichever is greater in the 1990s — and the surge test circuit condition that the total DC resistance including wiring, fixturing and the supply’s output impedance shall be a maximum of 1 Ω.
- Vishay. Tantalum Capacitors Frequently Asked Questions, document 40110. States that polymer capacitors use highly conductive polymer materials in place of multilayer MnO₂, giving much lower ESR, and that they do not have an ignition failure mode because of the lower oxygen content of the cathode.
- KEMET. Application Note for T599 Tantalum Polymer Capacitors in High Temperature Applications. Gives maximum series resistance from 25 mΩ to 150 mΩ at 100 kHz for the series, and a two-band voltage derating above 105 °C — 105 °C to 125 °C and 125 °C to 150 °C — with the derated voltage tabulated per rating.
- ECSS-Q-ST-30-11C Rev.2, 23 June 2021, Space product assurance — Derating — EEE components. Clause 5.4.2 holds the requirements for transient and surge conditions; the capacitor load ratios are in Tables 6-1 to 6-9 by family. Copyrighted; cited by clause, not reproduced.