How we can charge a 5V device monitoring the input voltage in real time

[testtest]

We have made a PCB using a MIC2876 to boost a 2.5V/4.5V 0.5 A input voltage to 5V output to be able to feed any device connected to two USB ports (USB-C and USB-A.)

We've also implemented a current sensor connected to an LED strip to monitor the incoming energy level.

We tested the unit by applying an input voltage in the range of 2.5V to 4.5V 0.5A but instead of boosting to 5V the output voltage is the same as the input - it fails its purpose.

We have tried to replace the IC and also built other PCBs but with the same result. We also tried to remove the current sensor to make sure it was not producing an excessive resistance, drawing all the incoming current but that didn't help, either.

Do you have any idea of what could cause the malfunction and do you have any suggestion to improve, if possible, the design?

[todorp]

I don't know it this is the problem but in the MIC2876 datasheet example (page 2) OUTS pin is connected directly to OUT pin...

[gamalot]

When you say "applying an input voltage in the range of 2.5V to 4.5V 0.5A", you mean that the power supply you are using has current limitation or overcurrent protection, right? Have you tried to increase the current limit?

[gamalot]

I don't know it this is the problem but in the MIC2876 datasheet example (page 2) OUTS pin is connected directly to OUT pin...

What you see is a reference design of a fixed output voltage models.

[testtest]

We try to increase the current range between 0.5A to 2A but the result is the same, the IC output is the same as the input

[Ice-Tea]

Points to the bypass mode being activated? "The MIC2876 automatically operates in bypass mode when the input voltage is higher than the target output voltage. In bypass mode, the NMOS is turned off while the PMOS is fully turned on to provide a very low impedance path from IN to OUT."

Have you checked the values of the feedback resistors? And the voltage at the FB node?

[testtest]

Hi Ice-Tea, thanks for your input but our feedback resistors are the same values ​​as in the datasheet

[Avelino Sampaio]

TestTest

Please, confirm for us if your chip code contains the letter "A". Example:

MIC2876-AYMT-T5

[testtest]

Hi Avelino

We used MIC2876-5.25YMT-TR

[nali]

So as already mentioned above looks like you're using the fixed output device with the reference design for the adjustable version.

Try removing R12 and replacing R11 with either an 0R or as low as possible value.

[Ice-Tea]

So, that's a non-adjustable version where you need to connect the FB pin directly to Vout (and that will regulate to 5.25V).

[testtest]

 now that I don't use the variable version I remove the resistors(R11 and R12) and short outs and out, but I still have the same problem

[DavidAlfa]

This might sound silly but happens: Are you sure the ic orientation is correct?
Could you post a a picture of the the pcb section of the MIC2876? (Design picture and real picture)

[testtest]

Hi David
I attach the images


[ Specified attachment is not available ]

[testtest]

And the other ones

[DavidAlfa]

I see weird copper traces in a lot of smd pads.
Is that an artifact caused by the gerber viewer?
Make some connectivity checks...

[nali]

My guess is it's oscillating, if it's doing anything at all. For example that looks like a *really* long GND path for the decoupling caps C1 & C8. If the artifacts David mentioned are actually there maybe it's an out-of-date copper pour?

[tooki]

I see weird copper traces in a lot of smd pads.
Is that an artifact caused by the gerber viewer?
Make some connectivity checks...

That looks like Altium (and not its Gerber viewer), in which case that’s bad news.

After my first pour screw up, I set Altium to always update pours.

[DavidAlfa]

Doesn't Altium run connectivity checks when exporting the gerbers to ensure everything is correct?
Should have spotted that huge mess up.

[testtest]

Thanks for your input but  before generating the gerber file, drc was done

[testtest]

if it can help here are the gerber files

[testtest]

This is a file where you can see the complete design that the machine sees when making the tile. Like this

[DavidAlfa]

Looks ok in the gerbers...

So have you:
- Ensured there're no shorts, specially under the QFN package?
- Shorted U1, pins 1-3,?
- Shorted R11?
- Connected a slightly higher voltage than Vmin, ex. 3V at VIN?

What are you getting at the output?
Have you checked SW with a scope to see if it's doing something?

[testtest]

Hi David

We have test all the part in your list and they seems fine
regarding the last 2 questions:
What are you getting at the output? the voltage output is exactly the same as the input, like is bypassing the IC
Have you checked SW with a scope to see if it's doing something? We are trying to find someone with an oscilloscope

[DavidAlfa]

Are you sure you got the fixed voltage version?
The datasheet isn't clear about the markings.

Vref is 0.9V in the adjustable version, it would explain why it bypasses the input.
Is the same happening when using the resistor feedback to OUTS pin?
Have you thoroughly checked their values?

Edit: The old Micrel datasheet is complete:
https://dtsheet.com/doc/1341016/mic2876---micrel

76H = Fixed 5.25Vout