Li-ion charger modeled after schematic on datasheet page 35.
DCIN 22 - 24V
Max charge current 20A.
How it works
With the battery at 6.3V, DCIN = 24V/3A from bench supply, Green LED comes on
INTVcc = 5.01V
FB = 1.42V
Enab = 1.49V
LB = 1.22V
Scope says it's 917kHz, with 232K on RT. (in-circuit with power off and caps drained it measures 146K.)
A DVM with a current shunt says there's no charging current.
The bench supply says it's supplying .011A
(With DCIN turned off, and an auto headlamp on the battery, the DVM says it draws 6.6A, so it can measure current.)
What do I need to look at to get it charging?
Something seems wrong here. If your DCIN is 24V, then you should be seeing that voltage as a maximum on the SW node and a greater voltage on TG. You mentioned the input supply is 24V/3A. With 20A charge current, perhaps you are hitting your power supply's current limit and that is drooping your input, possibly causing a restart? That would depend on where your input UVLO threshold is set.
I would also recommend that you reduce your charge current, just for now, to diagnose the issue. That way we can isolate problems with the circuit/layout from problems that show up at higher current levels. Perhaps that will get it charging for now and we can continue from there.
Turns out I omitted C29 from p35, so R22 was connected to ground. After adding C29, the top FET's quickly overheat, even with no battery connected.
Maybe multiple problems now.
Ok, worth checking that part number. Make sure it can handle the current that you expect. Vds(max) also needs to be greater than your input voltage.
Also be sure to calculate the values of the ITH RC for your specific application using the equation on page 28.
FET’s are all SiS468DN, four in parallel for top and four more for bottom, and two on INFET.
L is 6.6µH, R22 is 6.65K, and C29 is .01µF. R23 is 232K, so fsw should be 200kHz.
Are R22 and C29 close enough to the calculated values of 7392Ω and .0107µF?
The last test without a battery had red and green flashing alternately.
With the battery added green was solid, and after about 30 seconds one of the top FETs smoked.
Yes, I think those values are close enough.
Interesting that only one of the four parallel top FETs smoked? I wonder if you are having a gate drive issue with so much parallel gate capacitance. Perhaps they are not getting turned on fully and the partially-on resistance is causing the heat / blow-up.
Make sure your boost cap is sized appropriately according to the equation on page 27. If you have more than one board to play with, you may want to try pulling off some FETs so that you only have one on TG and one on BG. That should improve the gate driving if that's the issue here.
Each of these FETs can handle 30A so you should only need one of each. Did you just parallel them to spread the heat?