low battery hits to early 8x7135 driver

.

Hmm. Those drivers are pretty straight-forward for measuring voltage. They use a voltage divider - two resistors bridging power and ground with the center tap being connected to an ADC pin of the MCU. There’s some variation, especially with 5% resistors but I would expect it to be within 10%. 3.8V doesn’t sound good. When modding this thing, had you removed the driver?

What battery are you using and how old.
Some cheap batteries can’t even hold a voltage above 3v when a 3 amp load is applied.

Your missing the other voltage divider resistor. Unless its on the other side.
There’s a lot of variations of the amc7135 driver and I didn’t find your exact layout.
But it should have two resistors close to each other like in this version.

i would guess that you have a high impedance short with all the solder splatter, micro balls, residue, oxidation, dirt and debris. It’s easy to get a conductive ionic surface film growing with all that mix of chemicals.

Another possibility is a cracked ceramic capacitor or a solder joint that is changing characteristics with temperature during operation.

Well, after a little more searching I found your driver.
https://www.fasttech.com/products/1127403
It does appear to only have one resistor a 30D 200 kohms.
The one you have 154 is 150 kohms.
From Ohaya (once a active member here) comments it does have some type of LVP.
https://www.fasttech.com/forums/1127403/t/1001786/low-voltage-protection
Not sure how it reads the voltage but my guess to the problem is with kennybobby’s comments.
The board to me looks almost like a acid flux or acid core solder was used.
Try cleaning it with alcohol and see if that helps.

I’m no expert here but from what I gather, raising the value of R2 to 6 or 7 kohms should keep the voltage divider out put voltage high enough to make the mcu think the battery is still almost full.
Maybe someone else can chime in to verify.
It shouldn’t hurt anything in changing R2 values as the current thru the divider should be less than before with the total resistance increase.
It will however change the value of voltage of the LVP (if it has it) to a level that is unusable.

With the voltage divider, the mcu is reading the voltage thru R2 to ground. If you remove it then I assume it will read zero.

This might help a little. Voltage Divider Calculator
I was figuring this mathematcally to start with, the calculator is so much easier.
With a source of 3.4v (cut of voltage) the mcu should be looking to trip at anything below about .700 volts from the divider.
You can play around with changing values with R2 and R1 with different input voltages to adjust it however you would like.
A 50k for R1 and 15k for R2 would lower cut off to around 3v.