Replacing an Inductor

[ruthsarian]

I have a desktop LED lamp powered by a 12 volts DC 1 amp adapter. It also has a USB port on it to provide power. I found the power output from the USB port was really bad (looked like a sawtooth waveform on an oscilloscope). I opened it up and removed the power board. When I plugged it in I could hear some crackling, like there was a short somewhere. I found the source of the noise to be a 60uH radial leaded inductor. This inductor is inline with the positive lead that runs to the LEDs in the lamp.

I'm having a hard time finding a 60uH radial leaded inductor on places like Mouser and DigiKey. I do see 68uH inductors that have a similar form factor to the bad inductor. Would I be able to use a 68uH in place of the bad 60uH inductor in this application?

Thanks

[TheoB]

The inductor is probably fine. Even if it makes some noise. I guess it's there to regulate the current for the leds in an efficient way (step down converter). Since the leds work, the inductor should be fine as well. Only if you want to lower the noise you could consider to remove it by another one. 68uH is just as fine. Current rating is more important, so check that first. This has all little to do with the 5V regulator of the USB port that also should exist and does not work properly.

[ruthsarian]

There are two inductors on the board. After I removed one the noises had stopped so I thought that was it. Then I poked the other inductor and the noises started back up. I found if I pulled the second inductor away from the board the noise stopped and the 5v on the USB port looked good. I wrapped some copper tape around the inductor to shield it and pushed it back to the board and no more noise, the 5v remains clean, and the touch buttons to turn it on and control brightness, which were always flaky, now work great.

[ruthsarian]

Just a follow up to close this out. Even with the inductor 'shielded' I could still push on it and get see the power start to act odd. After a few hours the power problems returned so I ordered some replacement 68uH inductors. I've replaced the inductors and now the power is rock solid 5v no matter how much I poke and prod at the PCB.

Attached are a few pictures.

01: Now I'm very much the beginner hobbyist which is why you see the cheap pocket oscilloscope, but it's good enough to see that something was really wrong with the 5v rail on the USB port attached to this desk lamp.

02 & 03: The power PCB and the touch control PCB.

04: My failed attempt at the repair. The yellow inductor I had pulled first and the power problems went away for a little while. I thought the indcutor was bad and tore it apart to see what the guts of an inductor looked like. When the power problems returned I figured that inductor had been good and made a poor attempt at rewinding it. My cheap LCR meter says it lost about 10 uH, but it still worked! The center inductor was the problematic one. If I poked it you'd see the power dips demonstrated in the first picture. I thought maybe it was picking up some power off the PCB and tried wrapping it in copper tape then connecting it to the USB shield as a ground. That seemed to work. For a few hours. So it goes.

05: The inductors replaced and the 5v rail looking solid. I tried manipulating the inductors to see if it'd affect the 5v and it remained solid. So hopefully it's fixed.

[Yansi]

Veeery nice short circuit turn around the inductor, isn't it?

[TheoB]

The first picture shows the 5V USB power supply. Since this is made from 12V input it is probably created using a step down converter of which you replaced the coil. I suspect however that you have a problem with an output capacitor instead. What cap is present at the output? Most likely something like 10uF. Let's do some simple math. I observer the voltage drops by 3V in much less than 1ms. dV*10uF=30uC. To remove this 30uC from this cap in 1ms something needs to draw 30uC/1ms=30mA. And I think the fall time is even less than 1ms and thus the current should even be higher.
Perhaps the capacitor is not properly connected anymore. Try to replace it with a new one or check for bad solder joints. The inductor has only one failure mechanism that I can think of, and that is that there is an open. If that would occur, the output voltage would not show a ripple but it would just stay low.
The rising part of the waveform is about 2V/5ms. Or 2*10uC/5ms=4mA. I don't know the buck controller but 4mA seems like a low value for a USB output capable of driving 500mA. Have you measured the voltage at the input of the coil? I expect a 12Vpp input signal with a 5V/12V=42% duty cycle under load. Under no load it should be the same as the output of the coil (5V).
Another explanation for the big drops in voltage could be an output overvoltage protection that kicks in. And if the coil and cap would be OK, than there is an issue with the controller IC.