Canon binoculars IS - power supply fault

[jamarjan]

Good day to everyone!
I am trying to solve a fault in the stabilisation system power supply of Canon IS binoculars.
Battery power supply is based on the MB8300 regulator.
After inspection under the microscope, I found a damaged (corroded) via on the Vcc supply line just before the regulator IC.
After repairing the track and restoring power, unfortunately the circuit still does not work.

There is a sawtooth waveform on the OSC pin, PWM appears momentarily on the OUT pin, after which the circuit detects a short circuit on the output and shuts down the output.
A short circuit is detected even if the output path is disconnected (0 ohm resistor soldered out).
There is no short to ground on the output and all surrounding components have the correct values.
Does this mean that the circuit could be internally faulty or should I look for the fault elsewhere?

Link to MB3800 datasheet:
https://html.alldatasheet.com/html-pdf/902186/CYPRESS/MB3800/891/3/MB3800.html

[magic]

You say MB8400 and then post MB3800 datasheets. Is it a typo? Are these controllers equivalent?

How do you know that it "detects output short circuit"?

Does anything happen on the output node?

[jamarjan]

[small update]
I think im not getting a feedback signal on pin -in
Could this mean that Q108 is defective?
Multimeter measurements show no short circuit.
Does anyone have an idea what this component (PD 5 F4) could be?
Unfortunately with the help of https://smd.yooneed.one/ i am not able to identify it.

[jamarjan]

of course, MB8400 was a mistake  (will edit)

regarding short cirquit protection - take a look at the SPC pin waveform
the first rise is from soft start than it should settle low but instead it rises again and the OUT pin is pulled low

before shutting down the OUT pin square waveform looks ok (second pic)

at start up there is about 40mA current draw, then it drops to about 8 mA

[magic]

I would verify the feedback resistor going from output to IN-.

Then follow the signal path with a scope: the OUT pin, switching transistor collector, both sides of the diode, the IN- pin.


If I understand sections 10.3 and 7 correctly, the chip enters short circuit protection when it thinks that output voltage is too low for too long.
This may be bad transistor, bad diode, bad inductor, bad output capacitor, bad feedback resistor. Or bad chip, but hopefully not.

[fzabkar]

Does anyone have an idea what this component (PD 5 F4) could be?

http://markingcodes.com/search/c/pd

CPH5704, Sanyo, NPN Epitaxial Planar Silicon Transistor (15V, 3A) + Schottky Barrier Diode (15V, 1A), marking PD, package CPH5:

https://web.archive.org/web/20040625122721/http://prelcoparts.com:80/datasheets/Prelco%20Sanyo%20Data%20Sheets/CPH5704.PDF

[magic]

 
Yes, there has to be an NPN and a diode somewhere there. See datasheet section 11, application example.

If the output got shorted in the past, one of these components may be blown. This had happened to me in a similar (inverting) power converter on a hard disk. I replaced the whole transistor+diode combo with discrete components on appropriately bent legs.

[fzabkar]

If the diode is shorted or leaky, then the capacitor will look like a short when the transistor is PWM-ed.

[jamarjan]

HUGE THANKS fzabkar, you helped me a lot - owe you a beer or two ;-)

Schottky diode tests fine - Vf=155mV as does the NPN
Also tested the PNP on the output - also tests ok

[jamarjan]

Have to dig deeper i think.

After the CPH5704, before activating the MB8300 (push button) there is about 2,4V.
Activating the MB8300 results in voltage drop to 0V an some negative waveform - similar to MB8300 but reverse polarity.

[magic]

I don't understand what you are doing. Nothing here should be going below ground

Make sure that the circuit follows the datasheet example. Is the inductor connected to pin 3 (VCC)? The other end of the inductor goes to NPN collector and the diode.

Put one probe on SCP and the other on NPN collector, see what happens when the supply is turned on. Make sure that alligator clips are grounded correctly, to pin 6. Maybe there are different "grounds" there, or something, and you are measuring wrong one...

Does the circuit have any other connection with earth ground? I think not, you said it's battery powered. Not using a lab PSU to run it?

[jamarjan]

Indeed, I got a bit confused.
In the end I did, according to your recommendations, desolderd the coil and measured for 3v Vcc, and found nothing.
Unfortunately, the voltage supply trace to the coil is hidden inside the layers of the PCB.
Decided to run a wire from Vcc input on MB8300 to the coil and... IT WORKS !!!
Big thanks magic for commitment.
Cheers!

[magic]

For the record, I didn't recommend desoldering anything and there shouldn't be a need to. It's enough to check for continuity with DMM and/or measure voltage on the inductor's input side.

I'm also not 100% sure if the connection should be there, because in theory power to the inductor could be coming from a different source than the control chip's VCC. But if VCC goes straight to the battery, than the inductor should probably too...