(347) 552-2258 [email protected]

820-3332-A , orange light, but battery's not charging

Z

zarmin

New member
Hi!


I'm writing the whole story, may it contains additional valuable information.

One day, my macbook pro just switched off while working on it. Nothing trivial helped.
Fortunately I have a decent electronics lab at home.

PPBUS_G3H was missing, because Q7035 continuously shorted it to ground.
Some reflow with hot air and cleaning with IPA, around U7000 restored the PPBUS_G3H and machine turned on.

A few seconds later, the C7561 cap was started smoking, I've measured a few ohms of permanent resistance on that cap, so I've replaced it (I had only a 100 uF in a same package at home, so I've used that, instead of a 68 uF).

After that the machine started successfully, MacOS booted, etc.


The only issue is the battery.
My charging light turns to orange even there is no battery connected at all. When I connect it to the charger, it turns green immediately and after a second it changes to orange.
If I connect the battery, the machine doesn't recognize it (X battery icon), but also have an orange light.
I've done SMC / PRAM reset already of course.

But with every 8-10th startup the machine randomly recognizes the battery, displays its percent (29%), but can't charge it and tells me to replace it immediately, unfortunately if I remove the magsafe adapter, the machine turns off immediately.
I had a replacement battery also, which is also doing the same.


Thanks
 
Z

zarmin

New member
PPBUS_G3H with charge w/o battery: 12.51V which is somewhat lower than the expected 12.80V, but I think it's acceptable.
 
2informaticos

2informaticos

Administrator
Staff member
PPBUS_G3H is correct; no idea why you assume 12.8V!

Since 2013 all MacBooks turns MagSafe light to amber, even no battery connected.
Light goes green only if battery is 100% fully charged.

Post exact value of resistance between pins 17/18 and 27/28 of ISL6259.
Also post voltage at its pins 9 & 15.
Did you try original 85W charger?

Post R7080/81 and R7085/86 values measured onboard (ohm scale, no power).
 
J

JohnB8812

New member
Yep next thing to check is resistance between pins 17/18 and 27/28 of the ISL. If it is 12.51, your machine SHOULD be recognizing batteries
 
Z

zarmin

New member
1.
820-3332-A is a logic board for MBP retina 2012 mid, 15", 2.3 Ghz, 8 GB RAM (first rMBP)
So it's before 2013.
I don't know when should it show an amber/orange light.


2.
ISL 17, 18 pin, battery current sensing: 3.9 ohm
ISL 27, 28 pin, pwr supply current sensing: 21.3 ohm

by my calculations,

17, 18 should be:
2.2 + 0.005 + 0 = 2.205 ohm (+1.695 ohm diff)

27, 28 should be:
10 + 10 + 0.020 = 20.02 ohm (+1.28 ohm diff)

I used gold probes and negating the extra resistance on my probes, of course.

I've done some calculations about the possible resistance of PCB traces, and it should be between around 0.5 - 2 ohm, so these can be right values.

Also tested the intermediate values, so
btr current sensing:
R7051: 2.2 Ohm (should be, 2.2 ohm, 5%)
R7052: 0.1 Ohm (should be, 0 ohm, 5%)
R7050: 0.0 Ohm (should be, 0.005 ohm)

power supply current sensing:
R7021: 9.8 ohm (should be, 10 ohm, 5%)
R7022: 10.1 ohm (should be, 10 ohm, 5%)
R7020: 0.0 Ohm (should be, 0.020 ohm)

Aside from the measurement inaccuracy, everything in spec. I used Fluke 87V, currently I have no ready-to-use equipment that can measure low resistance precisely.


3.
I'm using my genuine apple 85W charger as always.


4.
I may have a theory, both of my batteries can be bad.
- The first battery had one cell which is swollen a bit, it was in my machine when it died (also had the replace the battery warning for days)
- The second was never used before, I've bought it to replace my old one, because of the warnings.
 
2informaticos

2informaticos

Administrator
Staff member
Check resistance to ground (diode mode, red probe to ground) for SCL/SDA lines of battery bus.
Be sure to have good solders on ISL6259.

Don't bother with exact resistors value; 10% error is OK for our readings, even for voltage.

Should be good to test known good battery; or test one of your batteries in other machine...
 
Last edited:
Z

zarmin

New member
5.
ISL pin 9 - GND: 0.298V
ISL pin 15 - GND: 0.042V
(/w power supply + battery)

R7080: 99.6k
R7081: 61.7k

R7085: 391.4k
R7086: 306.9k
 
Z

zarmin

New member
Check resistance to ground (diode mode, red probe to ground) for SCL/SDA lines of battery bus.
-> it's 0.530 in diode mode

Be sure to have good solders on ISL6259.
-> because at the first a quick reflow fixed my PPBUS_G3H voltage, it can be still an issue.
I'll remove the ISL6259, clean the pads and resolder it tomorrow, if no *easier* option will come until
 
2informaticos

2informaticos

Administrator
Staff member
Do you have .530 on both lines?
Less than 1% difference is required...

Check again R7085/86.
Invert probes and take readings in both ways.

Post voltage for pin 26/U7000.

"But with every 8-10th startup the machine randomly recognizes the battery"
This can mean cracked solders/traces, especially underneath SMC.
Pull it out and reball; with leaded balls, of course.
 
Z

zarmin

New member
I've measured it again
SDA: 0.526
SCL: 0.528
which is round 0.4% difference, so it's below 1%


R7086
CHGR_AGATE_DIV (+), PPDCIN_G3H (-): 307.2 k
CHGR_AGATE_DIV (-), PPDCIN_G3H (+): 325.4 k

R7085
CHGR_AGATE_DIV (+), CHGR_AGATE (-): 485.5 k
CHGR_AGATE_DIV (-), CHGR_AGATE (+): 391.2 k


26/U7000 voltage: 0.023 V


I'll try to get an another battery for testing, I'd like to spend my time on reballing the SMC on the last resort :)

Anyway thanks for your sustained and thorough help!
 
Z

zarmin

New member
After
- doing some reflow/cleaning around the SMC (not complete reballing)
- one of the pins of my battery connector was bent a little bit, so restored it
- doing PRAM / SMC reset
the machine reliably recognises my new battery.

My old battery is not recognized at all, but it was in a 'Replace now' state before and had degraded capacity, so it can be dead at all.

To check the details I've soldered probe wires to battery connectors i2c SDL/SCL lines (to the legs of D6950 diode pair) and checked it with my oscilloscope.
My old battery didn't produce any valid traffic on it, but my new one has valid i2c communication over it.


This is the output of System information with the new battery:

Battery Information:
Model Information:
Serial Number: D863105Q0TAF9Y2A5
Manufacturer: SMP
Device Name: bq20z451
Pack Lot Code: 0
PCB Lot Code: 0
Firmware Version: 511
Hardware Revision: 000a
Cell Revision: 1150
Charge Information:
Charge Remaining (mAh): 2579
Fully Charged: No
Charging: No
Full Charge Capacity (mAh): 8500
Health Information:
Cycle Count: 0
Condition: Replace Now
Battery Installed: Yes
Amperage (mA): 0
Voltage (mV): 11254

Can be any issue left on the logic board, that can cause this battery error? Or should it be the battery exclusively?
Because of the new year / weekend I can only get a new battery to test on next Wednesday.
 
2informaticos

2informaticos

Administrator
Staff member
If you get proper I2C communication on the battery bus, is time to test new one; or known good one, even older.
SMC should be OK, as is the master on that bus, and it starts the communication...
 
Z

zarmin

New member
Just wanted to tell, a new battery is working well. Both of my previous batteries were bad.
So my machine is working completely fine now.

Thanks for all of your help!
 
You must log in or register to reply here.
Top