LM5116: -15V @2 Amps

D5 passes a diode check with a Fluke.

Measures 9.1V on the cathode and 7.0V at the anode.

C17 is now 0.455uF ((2) 0.22uF and one 0.015uF stacked).

In the attached screenshot, Probe 1 is HO, Probe2 is LO, Probe3 is the output (-15V) (measures -0.241V with a Fluke, and Probe4 is UVLO.

There is a 1uF ceramic cap across R9/Ruv1.

Probe3 is on C12 (24Volts).

The power supply is at over 1 amp on the current limit.  It shows to be delivering 24V at 0.020A.  There is also a 5V and a 15V power supplies long with this -15V.

Touch a Fluke meter to DEMB this is the results.  The output Voltage is now -0.462V.  Probe3 is the output voltage.

Correction:

C17 is now 0.54uF ((2) 0.22uF and one 0.1uF stacked).

Here is a better view of HO and LO.

Hello James 

Please measure CS-CSG voltage and make it sure  V(CS-CSG) is below its current limit threshold. 

-EL

Channel 1: Output -0.447V Fluke

Channel 2: CS -0.434V Fluke

Channel 3: HO

Channel 4: LO

This is after probing DEMB/R13 with a Fluke.

Channel 2: CS at 1Volt resolution

Hello James 

I cannot read due to a noise. Can you zoom in V(CS-CSG) waveform ? 

-EL

I cannot read any useful information here. It is too noisy and I guess the DC offset is not calibrated. 

Channel 2: CS -0.434V Fluke

I assembled a second PCB with only the -15V populated.  It has the same problems.  I cut the traces to pin 13 CSG, ran a jumper wire from the pin and connected the other end to the -15V side of the 18mOhm sense resistor.  The next pic shows the circuit.

Channel 1: UVLO

Channel 2: HO

Channel 3: LO

Channel 4: Output -0.50V Fluke

I pulled the 1uF and 909 ohm resistor (R9) (the lower UVLO resistor) and replaced with a 1.5kohm.  I now get -1.26V out.

Channel 1: UVLO

Channel 2: HO

Channel 3: LO

Channel 4: +24V at C12

Changed the lower UVLO resistor to 4.7k ohms.  UVLO is now about 2 Volts which is above the 1.25V threshold.  The output is still too low.  Should be -15V.

Channel 1: UVLO

Channel 2: HO

Channel 3: LO

Channel 4: Output measure -1.25V with Fluke

Also to get the regulator started I have to touch the Fluke positive lead to the DEMB resistor.  This is the same as the first PCB.

Retook the last screenshot.  UVLO was not being displayed correctly.

Hello James

You have to figure out why the UVLO pin voltage was dropping. Since you confirmed the input power is stable, I think the UVLO voltage drop was caused by the hiccup mode operation. 

Also to get the regulator started I have to touch the Fluke positive lead to the DEMB resistor ==> Something totally wrong based on this information. I think you have to check all the connections match with your expectation. 

- EL