So I'm going to use the MAX8856 in my design and I have a 5800mAh 18650 battery.
The MAX8856 has a SYS output that cuts the output off at 3.5V. However, it seems that the 18650 would only be half discharged at that point. So what I want to do is put the system load in line with the BAT pin and the battery and use my own voltage monitor to be able to discharge the battery down to a lower voltage.
I've been speaking with the engineers, but is this possible/smart? Input? Thanks!
google "lithium ion battery discharge curve".
3.5v right around where the voltage falls off a cliff. You can damage the batteries by overdischarging them. Maybe you're getting it confused with a different battery chemistry?
Also, you do not have a 5800mAh 18650 battery. The ones from manufacturers who don't lie in their specs cap out at around 3000mAh last I saw (and cost 3 times what the liars' batteries do).
The ones higher than that, it's a crap shoot as to what the actual capacity is. I have some 4000mAh ones that have been measured at like 6-800mAh, and some 6000mAh ones that look pretty good - 2000+ mAh (but not 6000). I use the "6000mAh" ones - I needed enough of them that I couldn't afford to spring for good ones (as it was, I needed over $100 of them)
No, I'm aware of the chemistry. It's just the discharge curves I've seen don't fall off until ~3V and I've read that 3V is the minimum discharge voltage for 18650's. So 3.5V looks like it's right in the middle of the discharge cycle.
I'm also aware it's a capacity crap shoot. This is what the client wanted, so this is what I'm working with.
The discharge cycle is highly nonlinear though. It's not like alkaline batteries where it slowly falls off... Here's an example (pulled from google - there are lots more where that came from) - notice how 3.5v is right where the curve heads south. You don't have ~50% left, you have ~5% left at that point.
In any event, if you're spoofing the voltage monitoring, wouldn't it make more sense to get a simpler chip? sort of defeats the point of using a charge controller.
DrAzzy:
The discharge cycle is highly nonlinear though. It's not like alkaline batteries where it slowly falls off... Here's an example (pulled from google - there are lots more where that came from) - notice how 3.5v is right where the curve heads south. You don't have ~50% left, you have ~5% left at that point.
In any event, if you're spoofing the voltage monitoring, wouldn't it make more sense to get a simpler chip? sort of defeats the point of using a charge controller.
Could the difference be explained by: No Load Voltage (you) vs Loaded Voltage (OP) ?
DrAzzy:
In any event, if you're spoofing the voltage monitoring, wouldn't it make more sense to get a simpler chip? sort of defeats the point of using a charge controller.
The voltage monitoring isn't why I selected that chip. I went through almost 50 models and chose this one because of space constraints and functionality. It's not a big deal to change the kind of battery being used though. I can just as easily buy a battery that works better with the chip.
But I've been googling discharge curves for days. They all differ. I know most people don't Google things before posting, but I've been on this project a month and Google all day long. I wouldn't be posting if I didn't need clarification 
mrsummitville:
Could the difference be explained by: No Load Voltage (you) vs Loaded Voltage (OP) ?
Most curves I see show a discharge curve based off a load. Mine is ~1050mA.