beniceco/usb-cable-voltage-drop

USB cable voltage drop — gauge × length × current

A small lookup table for anyone modelling charging behaviour: how much voltage a USB cable loses on its power pair, depending on conductor gauge, cable length and charging current.

Every row is computed, not measured. That is the point: it gives a clean physical baseline you can compare real meter readings against. If your measured drop is far above the table, the extra loss is in the connectors or the cable is thinner than its label claims.

Columns

column	meaning
awg	conductor gauge of the VBUS/GND pair (20–28)
ohm_per_m	resistance of one conductor, annealed copper at 20 °C
length_m	cable length
current_a	charging current (5 A rows only for 20–22 AWG, which can carry it)
round_trip_ohm	2 × length × ohm_per_m — current flows out and back
drop_v	current × round_trip_ohm
v_at_device_5v / v_at_device_9v	voltage left at the phone for a 5 V or 9 V source
loss_w	power turned into heat in the cable, I² × R

How it was built

Resistance per metre comes from the standard American wire gauge table for solid annealed copper. Contact resistance of the plugs (typically 30–50 mΩ per contact pair) is deliberately left out so it can be added separately. The script is four lines of Python and is reproduced below.

awg = {20: 0.03331, 22: 0.05296, 24: 0.08422, 26: 0.1339, 28: 0.2129}
for a, r in awg.items():
    for L in (0.3, 0.5, 1, 1.5, 2, 3):
        for I in (1, 2, 3, 5):
            R = 2 * L * r; drop = I * R

Why it matters

A 3 m, 28 AWG cable at 3 A drops almost 3.8 V, which is why phones refuse to fast-charge over thin long cables at 5 V. Raising the bus voltage is the fix USB Power Delivery uses: at 9 V the same power needs about half the current. Above 3 A the USB-C specification also requires an e-marker chip in the cable, covered in the USB-IF USB-C documentation. A practical write-up on when a 5 A / 240 W cable is actually worth buying is on the CairoVolt blog.

Limitations

Temperature raises copper resistance by roughly 0.4 % per °C (see the temperature coefficient section of electrical resistivity), so a cable in a hot car loses more than the table says. Stranded conductors and copper-clad aluminium (common in cheap cables) are also worse than solid copper.