820-00840-A CD3215 short

[powerblackout]

Well I just removed some of the diodes that were reading 0L because I thought I was measuring them incorrectly. Measured both way in diode mode, 0L each way.

Can I temporarily remove the failing diodes just to see if I can get 20V, they say critical on them lol?

[mon2]

When the diode measures OL - each way then the diode is in open state = blown. It would not be of much use on the PCB to protect against transients. That is my understanding.

https://www.fluke.com/en-ca/learn/bl...to-test-diodes

Just the same, does not sound like they are the fault as to why you are not @ 20V. Unless the diodes are shorting the interfaced line to ground, these are not the root cause.

Check for any low resistance readings on the same rails which have these diodes. Any very low reading of resistance ?

[powerblackout]

ok, my brain stopped working for a second, i thought testing in diode mode meant 0L was bad, I see that now.

I am checking those rails for resistance now. I realized today that my donor board was charging the battery, before it was completely dead. Today I plugged the original board back into its case and PPBUS_G3H is present when the battery is connected!

On another note, when the battery is connected plugging in the AC power results in no display on the ammeter, all dead. Once I remove the battery the ammeter goes back to 5V like before. Not sure if this is helpful or not!

PP20V_USBC_XB_VBUS_CONN and PP20V_USBC_XA_VBUS_CONN both make the multimeter "flash" 0L as the autoranging flips between M and K ohms, I see it get to 35K or M ohms, then back to 0L. All of the rest of the lines (except for the ground pins) all have k or M ohms range on them.

[mon2]

Yes, this may mean that you have shorted mosfets.

You can measure the resistance across:

Source and drain
Source and gate
Gate and drain

For each mosfet. Checking for low resistance which means the mosfet is shorted.

Remove all power when you do this. Meter in 2k scale or similar.

Post your findings.

[powerblackout]

Ok I'll run through those. I couldn't help myself, I wired up the LCD and got everything put back in the case and started it on battery. backlight, display, and flashing file icon!

So just need to figure out why ammeter doesn't work... checking mosfets now

[powerblackout]

Alright I got the stencil and I reballed the CD3215 I got off of ebay. Turns out neither of them work, I took a CD3215 off of the donor board and it worked right away, both ports now go to 20V 0.01A, PPBUS_G3H is 0V so I need to figure out if U7000 is the issue now?

[powerblackout]

It appeared that U7000 was bad, I was not getting TBA_VDDA or TBA_VDDB those are now visible. After replacement the board still has charger at 20V 0.01A, but there is a 1ohm short to ground on PPBUS_G3H. I removed R7000 and the supply side (pin 1) has the short.

I got the U7000 replacement from rossmann's website, I really don't want to remove it to see if it is the cause because I don't have the means to reball and reinstall it. I started injecting 0.5V and moved up to 1.5V, the whole time the board was pulling 5A into PPBUS_G3H. There is something warm around F7000, possibly Q7040. I don't have a thermal camera but alcohol evaporates faster off of that one than the surrounding capacitors and Q7030. Any ideas?
I checked the CPU coils, all of them are 3-6ohms resistance, which from what I've read means it's not a CPU short.

I have removed and tested C7050 and C7051 which I thought where getting warm, but they did not resolve the short.

[mon2]

Apply flux on top of U7000.

Use high heat but low air pressure and keep moving the hit air tool in circles to cover the full top side of U7000.

Careful to not disturb any surrounding parts. If they move, then the air pressure is too high.

Have tweezers handy. As the hot air melts the hidden balls, use the tweezers to slightly nudge U7000 till the chip moves slightly. Eventually it will move with a gentle tap of the tweezers. Be patient. Then after this tap, the chip will bounce back in place to the natural pad site. Repeat for 2-3 taps while you keep hot air on top of U7000. Keep the hot air moving in circles.

This process will self align the balls that you cannot see and will remove any shorts.

Suspecting that you have a short underneath U7000.

When done, walk away and allow the logic board to cool down.

Then with no power, measure the resistance again that was showing low before on U7000. Proceed to test again.

[powerblackout]

Reflowed U7000 several times, no change. Removed chip, still have short

[mon2]

Removed chip, still have short

Where ?

[powerblackout]

PPBUS_G3H on the supply side of F7000
Removed Q7040 and short went away. Tested chip, 1ohm resistance between pin 1 (connects to F7000) and 6 (ground).

Now PPBUS_G3H is 5v lol

[mon2]

See attached. The 'X' mark the parts you have removed.

After the above, there is still a low resistance to ground on pin #1 of F7000 (fuse is removed).

Is this correct ?

If yes, then you have a shorted cap likely in the highlighted group of caps.

My bet will be on one of the 68uf caps.

Apply flux and remove one of the 68uf (C7050 / C7051) caps. Then check the resistance to ground again to see if the short is removed or not.

If not, proceed to remove the next 68uf cap. Post your update.

[powerblackout]

After removing Q7040 resistance jumped to 40mOhm. Replaced with chip off donor board now I get 5V on PPBUS_G3H. 20V 0.04A power connector

[mon2]

40mOhm

Small 'm' will be milli-ohm so this should be 'M' for Mega-ohms. That is good.

If the pad # 2 of the F7000 fuse is not with a low resistance, solder back F7000. Test again.

If still a no go...

Confirm all voltages on the CD3215 power rails (LDO).

Next, remove all power. Meter in DIODE mode.

Check each LDO rail in Diode mode:

Red meter probe to GROUND.

Black meter probe on rail to check. Post the voltage and diode readings for each rail & CD3215.

[powerblackout]

All LDOs were present on both CD3215, none of the rails had shorts. I decided to see what happens when running on battery and it killed the U7000, I guess I still had a short under it.

Replaced it with the original chip that still had some amount of balls on the bottom of it and it started working!

Thanks so much for your help!