Electronics > Repair
Noco GB70 Jump Pack Repair
demon1300:
--- Quote from: jerryk on March 09, 2024, 05:35:26 pm ---If your MC is shot the pack should not power up at all. No LED's, or partial working events will occur. That's because it has to latch the 3.3V LDO (IC6 labeled 7553-3) to it's on state. I would confirm that the output pin on IC6 is at 3.3V and not 3V or 2.9V. If it's output is being pulled down you have a short or load somewhere. The MC label is just a Holtek OEM number that applies to one of their existing MC's. The older GB70's were HT46R067 which is an OTP MC.
I would keep hunting but use caution. You still have your main 12V power wires connected and that is very dangerous if you are removing components to test things. For the heck of it check the four R005 resistors. If you can, hook up a battery to the main clamps and see it the relays come alive. Check all of your components around Q14 and Q15 for correct function. Then trace all three of the active wires in the relay connection at main board.
Edit to add: Also sniff around IC2 comparator circuit for faults. As I recall this is involved in the detection scheme of the voltage of the battery to be jumped.
Good Luck - Jerry
--- End quote ---
Thanks for the continued suggestions Jerry!
I've just checked IC6 and it's got 3.28v so that seems fine to me.
Using that MC data sheet that you gave me has definitely highlighted the problem. Pin 11 (PD1) which appears to drive the relay circuit has a short to ground (Pin 5 VSS). This clearly means there is a dodgy component in this circuit somewhere, i just can't seem to find it!
Tracing pin 11 back, it goes through 5 capacitors then to the boost switch. From there it looks to go over to R79. I've only tested this resistor in circuit, and get quite a different reading from the other GB70 i have here. On the good pack, it's reading around 563k ohm compared to 235k ohm. It looks to be marked with 1004 so am assuming 1m ohm? Not sure if this too far out of spec to be an issue?
The other issue i've found is that one of the pins on Q14 appears to have continuity to R74 (marked in blue on the diagram). The good pack i have does not have continuity here. I'd removed these components to see if there was anything wrong here and it looks like the problem remains. Does that maybe mean the board is toasted on an internal layer?
Multiple issues here, i'm just not experienced enough to seem to be able to connect the dots!!
jerryk:
The blue dot in your photo should only have continuity with PD1, pin 11 and that's it. I'm not aware of any internal layers on the GB70. The board I have is two sided but a bit older than yours. Since you have removed the only possible offending components (R74 and Q14) that I'm aware of there is a chance that you just might have an internal short in your MC. I haven't seen a partial MC failure in these Holteks but I suppose it's possible. Sure hope not. You can check Q15 for short's and diode check pins 2 and 3 but the short to ground at PD1 is not a good sign.
Jerry
grobius:
@demon1300, where you ever able to repair your unit? Exact same thing happend to me, a GB70 was fried with 24v and then to complete the day two GB150(yes 2, the level of stupitidy is off the charts!).
I've only started to look at the GB70, but from what I cas see the ht7533-2 LDO seems to be shot as it outputs same on Vout as Vin (12V). This leads me to think that the MC is also fried (and all hope is gone I suppose).
I'm courius about the zeners you changed. Do have the PCB labels and what values you used?
Noco should really have have added a protection circuit for this, but I guess that means that they would sell less units |O
Many thanks in advance!
jerryk:
Measure the voltage across C21. If it's in the 12V range it's more than likely game over for the MCU. Max input voltage per the datasheet is 6V. The input protection zener ZD2 is 5V1. ZD5 is 3V3.
Keep in mind as you explore this board that even with all JST plugs disconnected parts of the 12V rail of the board is still hot. The attached photo shows a safe way to work on the board by desoldering Bat+ and Bat- from the board and then making these connections removable via jumper wires.
Good Luck - Jerry
grobius:
Thanks Jerry, i will test this even though the odds are not in favour.
I wonder if anyone tried to dump the mcu fw and flash a new chip?
Maybe i Will be more lucky on the GB150s
Br Johan
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version