(347) 552-2258 [email protected]

820-00281 CD3215 madness

KC Micro Tronix

KC Micro Tronix

Member
No H2O on this board.

When testing the usb power would cycle on / off
I discovered one of the CD3215's was toast as a small hole was blown in it and it showed up quickly on thermal. Replaced it and it stopped the bounce but PP3v42 was a little low and the CD3215's were not talking to the charger and switching it to18v. Current draw was 0.08A at 5v. Thermal quickly identified the other CD3215 on the other side was also defective. Regardless of port used it was the one sucking PP3v42 down slightly to 2.9-3.0v . Replaced it and it and PP3V42 is now perfect. Current draw is 0.02 it just sits at 5V and does not switch to 18v. The CD3215C00's were replaced with the exact same part no the B03 versions.

Both defective CD3215's were on the same side of the board. Should I just shotgun the ones on the other side?

No shorts on the power rails. CPU / PCH are not shorted.

It has been about 6 months since I worked on these so I am just being a little lazy and looking for some guidance.


Thanks
Randall
 
Last edited:
S

SMMRepair

Member
In my experience, when a CD3215 has a hole in it, the CPU/PCH is toast. I've had 3 or 4 boards that had the CD3125's shorted so badly they burned a hole into the chip itself and fused the ball to the pad below it (I believe all shorted @3v3_g3h), and in all 3 cases, I resolved the short and CPU was killed. Look very hard at the CPU/PCH chipset; sometimes you can see the short/damage on the actual face/resin area. Looks like a tiny crack/brown spot...almost impossible to see without a microscope of course. Usually accompanied by a weird diode reading--i.e. .118 or similar. You'll sometimes notice the voltage is low and weird too, like 1.27v (so not STG).
 
2informaticos

2informaticos

Administrator
Staff member
CD3215 chips receive VBUS directly; and is 20V when working.
Depending of the way it gets burnt, input voltage can pass through (ms is sufficient) and sensitive chips from the board can be damaged; especially PCH/CPU and SMC.

You should check at least if all CD3215 chips generate 1V1_LDO.
 
KC Micro Tronix

KC Micro Tronix

Member
None of the CD3215's are producing 1V1_LDO

I tested the CPU and PCH rails for shorts and did not find any as stated before.
 
Last edited:
2informaticos

2informaticos

Administrator
Staff member
1V1 LDO must appear only on the CD3215 where the charger is connected.
Be sure to check that properly.
Compare diode mode readings for LDO outputs on all CD3215 chips.
You should have problems on USB-C areas, if cannot get 1V1 LDO.

BTW, no short on PCH/CPU rails cannot assure you the chips are good!
 
S

SMMRepair

Member
"BTW, no short on PCH/CPU rails cannot assure you the chips are good!"

Exactly! In fact, many times it WON'T show a short because it's not a total STG! Bet you have a dead CPU in this case.
 
KC Micro Tronix

KC Micro Tronix

Member
I hear you that the odds are low but I want to use this to learn on.

Thanks 2informaticos, I was measuring the 1v1 improperly. The 1v1 signal is present on the side that was not damaged. On the side I replaced the CD3215's neither one will show 1v1_LDO. Diode checking the 1v8 and 3v3 shows .5v normal readings. I compared them with the non damaged side.

So. Plugging into the non damaged side I get the 1v1 LDO signal on both CD3215's but the machine will still not swith to the 18v. On the damaged side where both CD3215's were replaced I do not get 1v1 LDO but the diode readings are the same as the non damaged side and appear normal. Both of the CD3215's on the damaged side were replaced.
 
Last edited:
2informaticos

2informaticos

Administrator
Staff member
If you have a board with dead CPU, or other "letal" problem, but gets 20V on USB-C ports, do not scrap it.
Use it to test CD3215 chips.
These chips can be reballed in 10-20 min with solder paste...
 
You must log in or register to reply here.
Top