I'm planning to buy an audio interface and it being exclusively bus powered is a requirement. However I also use condenser mics which require phantom power. Now I'm a bit confused about why some bus-powered interfaces require additional PSU for phantom power and some not. For example, audio interfaces lke Audient ID4 require connecting external power supply adapter when using phantom power, which seems logical. What I can't understand is how another USB 2.0 bus-powered interfaces like SSL2/SSL2+ provide phantom power on USB voltage alone - phantom power is 48v and USB 2.0 only provides 5v, if I'm not mistaken. The reason I'm asking is that I am choosing the audio interface and want to avoid any nasty surprises in the future. Any insights would be greatly appreciated.
Phantom power has very little current requirements, so a step-up voltage converter can easily provide +48V from USB power.
One reason not to do it (a choice some interfaces make) may be that designing a step-up converter in a manner where it doesn't interfere with high-sensitivity audio circuitry in the same enclosure is non-trivial. If you need phantom power for one input and another input is a high-impedance unbalanced instrument input, having the latter not pick up any converter whine in the range of the A/D converters requires careful design of internal shielding and operating frequencies.
Some work, some not. Creating 48V from 5V is done by a converter. Well-known technology but can be tricky to not induce noise in sensitive audio signals.
Back of the envelope calculation:
P48 should promise 10mA to the mic. If we calculate on the 10ma x 48V it comes out as 0.48W. Two mics and a voltage converter efficience of, say, 80%, the mics will require a power budget of 1,2W. Many mics draw less, say 4 mA, so we could might cheap out a bit on the power budget and most users will not notice.
Maximum current from USB2 is 0.5 A, times 5V gives 2.5W. In reality a bit less, as there are some cable losses.
This would leave a power budget of, say, 1.2W for the AD and DA and headphone amp, say 50mW output, and rest of the circuits. Can be done but may require either very good circuits or some "shortcuts".
It's possibly a measure of the unit's internal power-efficiency. If the interface needs most of the power for itself, then it doesn't have enough to spare for phantom too. I have never had any issues powering 2 high-end condensers from a single USB interface.
As mentioned in comments power transformers have no problem adapting voltage upwards as well as downwards, so long as the overall wattage/amperage is preserved.
You can even have 240v in your car, google 'power inverter'.
Others have described the theoretical basis -- that one can convert to a higher voltage with a corresponding drop in amperage. To go into the specifics for this hardware, I own a SSL 2+ and a USB-C power meter, so I'm able to produce actual measurements of the power draw with and without phantom power turned on. :)
These are instantanious snapshots -- if we wanted more solid numbers we'd want to actually record usage change over time, but eyeballing the numbers, they look pretty stable. Note that while I rounded the amperage and voltage numbers to two decimal points, the wattage numbers are based on the full instantaneous measurements.
- Running without phantom power turned on on either input: ~0.56A @ 4.96V (~2.79W)
- Running with phantom power turned on for one input: ~0.62A @ 4.95V (~3.05W)
- Running with phantom power turned on for both inputs: ~0.65A @ 4.94V (~3.22W)
- Actively recording both inputs with phantom power on: ~0.65A @ 4.93V (~3.20W)
Going from no use of phantom power at all to phantom power on one input, we have an increase in power usage of about 0.26W; adding the second input, that increases by an additional 0.17W; and whether we're actively recording or just monitoring appears to have no impact on power draw.
The USB 2.0 battery charging extension allows 1.5A@5V on a standard USB-A port; if your laptop supports this extension -- or you upgrade to a laptop with USB 3.0 support that supports the 4.5W "high-power SuperSpeed" profile -- you'll be more than fine. On the other hand, if your laptop tops out at the .5A@5V "high-power device" profile, you might have trouble.
This test was run with two microphones plugged in: An Aston Origin on the first input, and a Movo LV8-C on the second. The power source (relevant to the voltage drop numbers) was a Dell XPS 13 9370 plugged into its preferred 20V upstream power supply.
The problem is not the voltage (it can be converted from 5 to 48 inside the interface) but the total amount of power available. If the current supplied by the computer to the USB is enough, the thing may work. But you're right to be concerned, I think this is a borderline situation, and although things may work with just USB power, it could be wise to get an interface with an external power unit, to be sure you'll never have this problem.