USB to 3v 100mA. What did I do wrong?

[benjman]

GOAL: Power a 2x AAA battery powered headset with a USB port

METHOD: 3x diode in series

RESULT: Failure

Detailed Methods:

Connected three diodes head to foot. Connected USB 2.0 (side with USB to computer) red wire to 1st diode @ the opposite side of the band. Connected an empty red wire to 4th diode (also tried 3 diodes with no change in success). Connected black ground wires. Connected wires to headset (first red to positive battery terminal wire, then switched). None of these allowed the device to power on when USB was plugged into computer. Using a voltameter, I found there was a circuit when USB plugged into computer. For whatever reason, in DC mV I saw a 200 mV signal. Please help, Thanks

 

[meBigGuy]

USB may not stay powered up if there is no device connected. And, some may not stay connected if the device does not respond. One can signal device connection to USB (as a full speed device) by connecting a 1.5K resistor to 3.3V and connecting it to the DP pin.

Also, drawing too much current can cause shutdown. The max USB load is 44 ohms in parallel with 10uf (1 unit load) which is 100ma at 4.4 volts (the lowest V+ can be). I suggest you play a bit with a constant load and look at what the V+ line does over time. Does it initially produce 5V and then shut down? Or, are you just drawing > 100ma and it is shutting down.

There are some tricks you can play with the USB dead battery provision to get a modern USB port to give you power for a while. But the only way to get reliable power or more that 100ma is to enumerate.

I've built lots of devices that charge or run from USB, but I've always had fully implemented devices that communicate with the host and could request 500ma. I haven't played around with what Host ports do with "illegal" devices.

 

[benjman]

hmm so with the multimeter connected before i plug in the usb, i get a 4 V output. I have 4x 1n4001 diodes in series so i don't know why its only a 1V decrease? anyway, Ill try using this with the headset power held down before plugging in the usb to see if that works. maybe 4V is too much for the 3V headset??

UPDATE: hmm didnt work. I am stuck now. would decreasing the 4V to 3V fix the problem?

 

[meBigGuy]

No idea.

Are you sure the two batteries are in series? Sometimes they put batteries in parallel and have a switching converter.

The 1N4001 will drop what you expect if there is enough current being drawn. I don't know whether you are reading the 4V with no load (just the multimeter). Have the diodes drive a 47 ohm load and measure across that, with no headphones. That at least let's you determine what the USB will do for well behaved loads.

I have no idea of the current requirements of your headphones or what kind of power system they implement. It could be that 100ma isn't enough.