RIPPLE PACK CALCULATOR

The caps at the ESC, sized on a real run: launch, throttle sweep, and heat — for drag racing or speed runs

Cap values are RATED from the maker's data sheet unless tagged EST. The E6 Max doesn't publish its switching frequency or its onboard caps, so frequency, launch profile and ESC cap share are EST. Confirm with a logged pass.

New to this? Most people only change cells in series, launch amps, pass length, and the cap preset. Leave the other numbers unless you have a logged pass — they are EST starting points. Every field has a ? that says what it means. Starting numbers are example assumptions, not measured EVUL data.

Drag racing or a speed run?

This calculator is for either one. The job of the cap bank is the same: it sits at the ESC and feeds the high-frequency current that long battery leads cannot.

Drag racing: The starting numbers are built around a short pass (3 seconds) and a hard launch. If that matches your car, you can leave them and just set cells in series, launch amps, and the cap preset.

Speed runs: Change Pass length to how many seconds the motor is under power. A speed run is usually much longer than a drag pass, and that extra time puts more heat into the caps. Set launch peak to your ESC limit. Then open More options and set the after-launch current and mid-pass power from a logged run if you have one; if you don't, the defaults are EST starting points, not a speed-run measurement. Check heat per pass and the heat-stacking table if you make several runs without a long cool-down.

Run conditions

Cells in series S

How many cells are stacked to make pack voltage. 8S is eight cells in series. This is the S in 8S2P, not how many packs you run in parallel.

Charge per cell V

Voltage of each cell at the start of the pass, usually a full charge. 4.20 V is a typical LiPo full-charge figure. Pack voltage used here is S × this number.

Launch peak, motor amps A · ESC limit

The highest motor current at launch, usually your ESC's current limit. On an E6 Max that is often 550 A. This spike only lasts milliseconds; it is not the current for the whole pass.

Pass length s

How long the motor is under power, from launch to the end of the run. The starting number is 3 seconds, a typical drag pass. For a speed run, put in your actual run time — longer runs put more heat into the caps.

PWM frequency kHz · E6 Max unpublished, 20 EST

How many times per second the ESC switches the motor. Faster switching usually means less ripple voltage for the same cap bank. The E6 Max does not publish this number, so 20 kHz is EST. Confirm with a logged pass or the maker if you have it.

Peak decay time ms · EST

How quickly that launch spike falls off after the car hooks up, in milliseconds. The default 50 ms is an EST starting point, not a logged number from your car. A bigger number means the spike lasts longer and the caps work harder.

Traction-limited motor amps A · EST

The current the motor pulls after launch, once the tires have grip, not the ESC's peak limit. The default 300 A is EST. If you have a logged pass, use the current you see after the first tenth of a second.

Pack power, mid-pass W · EST

About how many watts the motor is making in the middle of the pass. This caps current once throttle is up. The default 5000 W is EST. Leave it unless you have a better number from a log.

ESC onboard caps take this share of ripple 0 %

How much of the ripple current the ESC's own input capacitors already absorb, so your add-on pack does not have to. The E6 Max does not publish this, so the slider starts at 0% — the worst case. Leave it at 0% unless you have logged it.

0 % is the worst case. How much the E6 Max's own input caps absorb is not published EST; leave at 0 % unless you have logged it.

Thermal limit per pass 10 °C

How many degrees Celsius you will allow each cap to warm during one pass, with no cooling between runs. The default is 10 °C. This is a comfort limit you set, not a data-sheet number.

Ripple capacitor

Preset

Pick a capacitor from the list and the capacitance, ESR, ripple rating, mass and voltage fill in from the maker's data sheet (RATED). Choose a custom option if you want to type your own numbers.

Capacitance µF

How much charge one cap can store, in microfarads. More µF lowers the part of ripple that comes from charging and discharging. On these low-ESR caps, ESR often matters more than extra µF.

ESR mΩ @100 kHz

Equivalent series resistance — the cap's internal resistance, in milliohms, usually given at 100 kHz on the data sheet. Lower ESR means less ripple voltage and less heat. N caps in parallel divide this by N.

Rated ripple A rms

The maker's continuous ripple-current rating for one cap, in amps rms (RATED). A 3-second pass is a heat problem, not a lifetime problem, so we also check temperature rise per pass.

Mass g

The weight of one cap, in grams. Used to estimate how many degrees it warms in a pass. Heavier caps soak up the same heat with less temperature rise (EST).

Voltage rating V

The maximum voltage the cap is designed for. It must stay above your full pack voltage (S × charge per cell). A 63 V cap on an 8S pack at 4.20 V/cell (33.6 V) has comfortable margin.

Caps in bank N

How many of this cap you put in parallel on the board. More caps share the ripple current, cut the effective ESR, and add weight. Use recommended / minimum / full-rated to fill this from the model.

Ripple at the ESC input

Three ways to size the bank

Ripple vs PWM frequency, this bank

Heat stacking, no cooling between passes

Heat per pass = (pass rms per cap)²·ESR·t ÷ (mass × 1 J/g·K EST). Ignores cooling (conservative) and core hot-spot (optimistic). Continuous ripple ratings are for thousands of hours at 105 °C; a 3-second pass is a thermal problem, not a lifetime problem.

Battery sag?

Ripple caps don't fix sag. Sag is pack resistance × amps; the fix is more cells in parallel. Use the How Much Battery? calculator.

How we chose these capacitors

The calculator does not pick a capacitor for you. You pick a part (or Custom), and it figures how many you need for a real pass: launch spike, throttle sweep, and heat.

The preset dropdown follows cells in series. Change S and the list swaps to parts sized for that pack voltage class: 2S–4S favor 35 V cans, 5S–6S favor 50 V, and 7S+ favor 63 V (with any 50 V option marked as thinner margin). Custom stays put if you were already typing your own numbers.

Shortlists were scored on November 6, 2026 with the same rules the calculator uses at its starting numbers for that class (4S / 6S / 8S at 4.20 V/cell, 550 A launch, and the EST starting points on the page).

Through-hole, axial, or solder-pin cans only. Surface-mount parts were left off. We preferred a comfortable voltage margin (~10 V leftover). Ranked by how many caps the calculator recommends (fewer is better), then by bank ESR. “Recommended” means sustained ripple ≤5% of pack voltage, launch spike ≤10%, and heat within the 10 °C per-pass limit. If you change your run, the count changes.

Stock moves around, so look the part number up at a parts house before you order. Mass figures in the model are EST from can size unless a maker publishes weight.

4S class (scored at 4S · 16.8 V)

#	Part	µF	V	ESR	Ripple A	Can mm	Rec N	Note
1	TDK B40620B7148M001	1400	35	4 mΩ @20 kHz RATED	11.8	16×25 axial	5	4S dropdown default · hybrid · very low ESR
2	Nichicon UHW1V392MHD	3900	35	10 mΩ @100 kHz RATED	4.20	16×35.5	5	Radial TH · ties on count, higher ESR
3	Nichicon UHW1V332MHD	3300	35	11 mΩ @100 kHz RATED	4.00	16×30.5	6
4	Rubycon 35ZLH2200MEFC16X25	2200	35	13 mΩ @100 kHz RATED	3.63	16×25	7	Common ZLH radial
5	Nichicon UHW1V222MHD6	2200	35	13 mΩ @100 kHz RATED	3.81	16×25	7

6S class (scored at 6S · 25.2 V)

#	Part	µF	V	ESR @100 kHz	Ripple A	Can mm	Rec N	Note
1	Rubycon 50ZLJ2700M16X40	2700	50	14 mΩ RATED	4.80	16×40	5	6S dropdown default
2	Rubycon 50ZLJ2200M16X31.5	2200	50	18 mΩ RATED	4.10	16×31.5	6
3	Nichicon UHW1H472MHD	4700	50	20 mΩ RATED	3.86	~16×35.5	6
4	Panasonic EEU-FR1H102	1000	50	16 mΩ RATED	3.32	16×25	7
5	Rubycon 50ZLH1000MEFC16X25	1000	50	21 mΩ RATED	3.01	16×25	8

8S class (scored at 8S · 33.6 V)

#	Part	µF	V	ESR @100 kHz	Ripple A	Can mm	Rec N	Note
1	Rubycon 63ZLJ2200M16X40	2200	63	15 mΩ RATED	4.10	16×40	4	Electrically best; taller can; check stock
2	Rubycon 63ZLJ2200M18X35.5	2200	63	17 mΩ RATED	3.90	18×35.5	5	Same family, slightly wider/shorter
3	Rubycon 63ZLH2200MEFC18X40	2200	63	18 mΩ RATED	3.67	18×40	5
4	Nichicon UHW1H472MHD	4700	50	20 mΩ RATED	3.86	~16×35.5	5	50 V, thinner 8S margin
5	Nichicon UHW1J222MHD	2200	63	20 mΩ RATED	3.86	18×40	5	8S dropdown default
6	Panasonic EEU-FR1H102	1000	50	16 mΩ RATED	3.32	16×25	6	50 V, thinner 8S margin
7	Rubycon 50ZLH1000MEFC16X25	1000	50	21 mΩ RATED	3.01	16×25	6	50 V, thinner 8S margin
8	Rubycon 63ZLH1500MEFC16X35.5	1500	63	21 mΩ RATED	3.04	16×35.5	6
9	Nichicon UHW1J152MHD	1500	63	23 mΩ RATED	3.10	16×35.5	6
10	Panasonic EEU-FR1J821	820	63	24 mΩ RATED	2.99	16×25	8	Same 16×25 as old default

The previous 8S default Rubycon 63ZLH820MEFC16X25 (recommended count 9 at 8S starting numbers) remains in the 8S dropdown. Mass values used in the model are EST unless a maker publishes weight.

The math

Throttle rises over the pass: D(t) = 0.05 → 0.95 (EST). Motor amps: I(t) = min( I_hold + (I_peak − I_hold)·e^(−t/τ) , P/(0.85·D·V) ) Cap ripple current = I·√(D(1−D)) (zero at D≈0 and D≈1, max at D=0.5) · ΔV_pp = I·D(1−D)/(f·C·N) + I·ESR/N Heat: I_rms over the whole pass, per cap, ²·ESR·t / (m·c) · Sized on sustained ripple (after 0.2 s), spike only checked

Why not 79 caps? The old sheet put the full 550 A launch peak at 50 % throttle for the whole 3 s. That never happens: 550 A is a launch spike at low throttle, and by the time throttle reaches 50 % the motor is pulling a few hundred amps, then ripple falls away again near full throttle. Real ripple heating over a pass is about 116 A rms total, not 275 A. The bank is there to feed the high-frequency current the battery leads (≈1 nH/mm) can't, right at the ESC. Commercial packs agree: Castle CapPack 4 × 220 µF 50 V, Castle 8S 3 × 560 µF 35 V, Team Brood Speedrun 5 × 2400 µF 63 V. Not modelled: the E6 Max's own onboard caps (use the slider, EST), ESR rising at 20 kHz vs the 100 kHz datasheet value, lead inductance, ESC current limiting.

Not modelled: lead inductance (≈1 nH per mm of loop; 40 nH is 5 mΩ at 20 kHz, so the board still bolts to the ESC), LiPo polarization and recovery, the ESC's cycle-by-cycle current limit, and ESR drift with temperature.