USB Power Supplies
General Background
TBD
Test Kit/Method
|
As can be seen above, testing was performed with a breadboard configured to allow a number of resistors to be switched in parallel across the supply under test via a DIL switch package. Four (56Ω) resistors were mounted on the top-side of the board and the other four (47Ω) were mounted underneath. Virtually all of the components were items purchased, maybe, 20-30 years ago - the exception being the USB socket. Ideally I would have liked to use eight resistors of the same value but I didn't have such "in stock". The calculated and measured parallel resistances are shown in the table below - for a (constant) 5V supply this leads to expected nominal load currents of 100-800mA in 100mA steps.
Switches "On" |
Calculated Ω |
Measured Ω |
1xxxxxxx |
47.0 |
45.9 |
12xxxxxx |
25.6 |
25.4 |
123xxxxx |
16.6 |
16.4 |
1234xxxx |
12.8 |
12.7 |
12345xxx |
10.0 |
10.0 |
123456xx |
8.5 |
8.6 |
1234567x |
7.2 |
7.3 |
12345678 |
6.4 |
6.4 |
Tested Devices
|
Test Results
(See also OEM Hub PSU, DVE Hub PSU, iLuv i108 5V 2A PSU, Powered USB Hubs and Other Tested PSU's.)
|
|
|
|
Good Enough for the Pi?
The power budget of a model B Pi is associated with "self-current" of 700mA plus 200-300mA for attached USB devices. Hence a 5 Watt (5V at 1000mA) PSU should be fine. The "FX" and "TCB" devices both meet that specification and the samples measured confirm their suitability. Whilst the Sony Ericsson charger has a "below spec." maximum current it is the most stable of the devices tested and has been used to power a first generation model B Pi running "headless" ie. without monitor, keyboard or mouse but with one or two USB flash drives plugged in. This is the Pi shown on the front page and on which this wiki is normally running. Whilst the "Generic" devices were specified as "5V 1000mA" they do not strictly meet this, or rather the "1000mA" appears to be their short-circuit current limit, and hence these are not recommended.