No bench test has been run yet. Until results are published, the calculator and battery pages use a stated planning assumption: Planning assumption: Our stated assumption, used until bench results replace it about 80% of a power bank’s label Wh, 85% of a power station’s and 90% of a LiFePO4 battery’s is usable. We won’t name a “best” brand before we’ve measured it.
The test list
| Battery | Port | Load | Label Wh | Delivered Wh | Status |
|---|---|---|---|---|---|
| USB-C PD power bank, about 99 Wh, 100 W single port (model chosen at test time) | USB-C PD 20 V | 40 W | — | — | Not run yet |
| A second 99 Wh, 100 W USB-C PD power bank from a different maker | USB-C PD 20 V | 40 W | — | — | Not run yet |
| A 65 W USB-C power bank (to confirm it can't hold the 20 V / 5 A profile) | USB-C PD 20 V | 40 W | — | — | Not run yet |
| Small power station, about 256 Wh | USB-C PD 100 W | 40 W | — | — | Not run yet |
| Same small station, AC outlet | AC (built-in inverter) | 40 W | — | — | Not run yet |
| Mid power station, about 512 Wh | 12 V DC | 40 W | — | — | Not run yet |
| 12 V 100 Ah LiFePO4 battery | 12 V DC (fused) | 40 W | — | — | Not run yet |
| The owner's home Starlink Standard dish + router on a UPS: watts at idle, in use, and the switchover in a pulled-plug test | AC (UPS) | — | — | — | Not run yet |
Why test batteries at 40 W?
The top of Starlink’s average range for the Mini is 40 W Official spec: From a Starlink specification sheet or help article. Checked Oct 5, 2026. A constant 40 W load through a 20 V USB-C PD trigger makes a battery deliver energy the way it would to a Mini on USB-C, without needing a Mini. That matters because a battery’s converter is less efficient at some loads than others, and some power banks can’t hold 20 V at high current at all.
Equipment
- USB-C PD trigger board set to 20 V: Makes a power bank deliver the same 20 V PD profile the Mini asks for.
- DC electronic load (constant-power mode) with logging: Draws a fixed number of watts until the battery cuts off, and logs volts, amps and watt-hours.
- USB-C inline meter (V, A, W, Wh): Cross-checks the load's reading at the port.
- Plug-in AC power meter: For AC outlets through an inverter, and for the owner's home Standard dish and router on a UPS.
Meter models and their stated accuracy will be listed here with the first result.
Protocol
- Charge the battery to 100% and let it rest 2 hours at room temperature (68–77 °F). Record the temperature.
- Set the load to a constant 40 W, the top of Starlink's average range for the Mini (Mini on USB-C or 12 V; a second run at 25 W where time allows).
- Log volts, amps and watt-hours every 10 seconds until the battery cuts off.
- Delivered Wh = the watt-hours logged at cutoff. Delivery ratio = delivered Wh ÷ label Wh.
- Repeat once. Publish both runs, the photo of the setup, the meter models and the raw CSV.
- For AC: same load through the battery's AC outlet into the plug-in meter; the difference from the DC run is the inverter cost.
What this bench can’t tell you
- This tests what the battery delivers at the Mini's input profile. It is not a test of a Mini: the owner doesn't have one.
- One sample per model, at room temperature. Cold cuts capacity; we'll say so where it matters.
- Meter accuracy is stated with each result.
We buy the batteries we test. If a maker ever sends one, the result says so and says whether we kept it. Whether a brand pays commissions never changes a result or a pick (editorial policy).