...
note there is a 4A current limit on the input power supply, because below this cannot overcome the ICLs and bulk caps
Test number: 1,2,3,4,5,6,7,8,9,10,11,12
input voltage: 90.6, 90.6, 90.6, 90.5, 130V, 130, 130, 130, 143, 143, 143, 143
...
voltage after ICLs and fuse:85.5, 86.6, 87.3, 87.6, 125.2V(3.82 drop on ICL), 126.3(2.9 icl 70C other is 31C), 127.4(2.2 icl 50C and 31C), 126.8 (2.3V icl 107C and 37C), 138.2 (4.3V icl 33C and 31C), 137.8V (4.0V icl 55C and 40C), 139V (3.0V icl 81C and 37C), 139.7 (2.4 icl 97C and 37C)
Discharge circuit
From visual testing, the DCDC fault led turns on when the input has been removed and the load still exists until the capacitor is drained in which it turns off. This lasts up to 1 second for any load or supply, proving discharge works.
At 74V discharge is enabled (this is because the transistor is no longer fully on)
Transient Analysis
states:
...
Ripple Output at various conditions
2A out, 90V in (84.6 after fuse), 100Hz, 580mV (low frequency, probably not from switching noise)
6A out, 90V in (84.6 after fuse), 100Hz, 640mV (low frequency, probably not from switching noise)
No change for higher input voltage
16.5kHz and 4.6Vpp for ripple when the non-noisy load was used during transient analysis
Audible noise
Heating
at 90V input 1.73 A input and 10A output, ntc3 is 90c and ntc2 is 40 → vdrop was 2.4V
...
at 90V input 2.69A output 16A, ntc3 was 150C, ntc2 was 30C
Fault Mode
Overcurrent starts at 19A out
Undervoltage lower threshold 40V upper threshold at 44V
From the above testing, the following ICL has been specced for use as a single ICL in series. Note that the decision to use a single ICL in series comes from the fact that small differences in resistance for the ICLs in parallel causes one to get much hotter than the other.
B57236S0809M000
B57236S0120M000