EEVblog Electronics Community Forum
General => General Technical Chat => Topic started by: hamster_nz on October 30, 2019, 10:37:10 pm
-
I had the displeasure of looking at a the 5V rail of USB power adapters, while trying to track down the source of noise in a project I'm building. The only one that was anything like a good supply for evaluating audio projects was the one that came with my ChromeCast Audio.
I must say, the level of noise and switching artifacts varied really wildly. What surprised me most was a "Power Bank" that jumped up and down 0.1V for 20ms at a time... maybe it was probing to see if it needed some high current mode.
-
The power bank is probably probing for a load yeah, so it can shut down the output if nothing is connected. That or its attempting cable compensation. It doesn't seem like pulse skipping.
Anyway, USB adapters are focused on efficiency and quiescent current, not noise. I would probaby never use the output directly in any serious project. You could always drop the rail to 4 or 4.5V with LDO, etc.
-
I was looking at Pk-Pk +/-10mV after a lm1117-3.3V LDO, with lots of high frwquencies in it, that was leaking into the ADC.
Sigh... But it is all learning experiences :)
-
I was looking at Pk-Pk +/-10mV after a lm1117-3.3V LDO, with lots of high frequencies in it, that was leaking into the ADC.
Sigh... But it is all learning experiences :)
Thats surprising to see 10mV. The 1117 looks like it tapers off at about 20kHz or so, then noise will get through. Could try some ferrites for the HF, one on input one output.
Maybe daves capacitance multiplier is worth playing with as well: https://www.eevblog.com/forum/blog/eevblog-1116-the-capacitance-multiplier/ (https://www.eevblog.com/forum/blog/eevblog-1116-the-capacitance-multiplier/) I haven't tried, would be cool to measure its frequency response compared to the LDO.