I have a solar panel battery charger project, but have issues with the voltage dropping from the output of the buck converter (LM22680).

Setup:
Bench power supply primarily set to 12V. Current limit set to 100mA. Used as PWR_IN
J105 Shunted

Observations:
1) With no load (J108 left open), the output of the LM22680 is a clean output voltage
2) With a 1k resistor as a load, the output of the LM22680 is a clean output voltage
3) Connecting the output from LM22680 to the battery charging chip (LTC4055) via J108, the output from the LM22680 drops (~0.7 - 1.2V) depending on the input voltage.
- If PWR_IN is set to 12V, the mean voltage on VIN_+5V is ~3.044V with a current draw of ~8mA ( Waveform #001 )
- If PWR_IN is set to 8V, the mean voltage on VIN_+5V is ~4.168V with a current draw of ~16mA ( Waveform #002 )
4) LEDs D107, D108 (CHRG, ACPR) flicker ON/OFF even though the voltage on WALL is set above the threshold voltage of 1V. I've changed resistors R111, R112 so that the voltage on this pin is ~3V so that any drop in voltage on VIN_+5V will still keep the LTC4055 ON and Charging. But it doesn't seem to work.
5) Bypassing the buck converter circuit and applying voltage to the LTC4055 (J108, pin 2) through the Benchtop power supply (5V), the battery charging circuit works fine and the voltages are stable.

Questions:
1) I've set the output voltage for the LM22680 to be ~5V (via R102, R103), why does the output voltage from LM22680 change drastically depending on the power supply input, even though it is well above the set output voltage of ~5V? Is this an efficiency issue? A chip conflict issue with the LTC4055 chip switching ON/OFF?
2) Why is there a pulse (noise) every ~1ms? Is this normal for buck converters under load?
3) How can I cleanup the VIN_+5V voltage signal? Adding a 1000uF or 4700uF polycap didn't seem to have any effect

Schematic in attachments

Waveform #001


Waveform #002

 

Could you increase the current limit of the bench PSU and see what happens then?

 

Could you increase the current limit of the bench PSU and see what happens then?

No difference. I increased the current limit to 1A, but at 12V, it's still only using up ~8mA. At 8V, it's using ~14mA.

 

Could you explain why the enable pin is connected to the PWR_FLAG label? That part of the circuit seems a bit suspicious to me, but I could be wrong.
Also your resistor values for the enable pin seem to be a bit dangerous:

 

Could you explain why the enable pin is connected to the PWR_FLAG label? That part of the circuit seems a bit suspicious to me, but I could be wrong.
Also your resistor values for the enable pin seem to be a bit dangerous:
View attachment 218500

For the LM22680 VIN pin, the PWR_FLAG is connected. It's just the text is close to the EN pin.
I did have to change the resistor values for the EN pin, R104, R105 to R104->100K, R105 ->34.8K so as not to exceed 6V. I should have updated the schematics.

 

Sorry I didn't get that.

In your battery charging circuit there is also a PWR_FLAG label. Are these points electrically connected? That would imply that your VIN is directly connected to your VIN_+5V. Doesn't make any sense, but I have to ask.

 

Sorry I didn't get that.

In your battery charging circuit there is also a PWR_FLAG label. Are these points electrically connected? That would imply that your VIN is directly connected to your VIN_+5V. Doesn't make any sense, but I have to ask.

No they are not electrically connected. The PWR_FLAG is just a symbol to mark a pin/net as a power source.

 

Here are some waveform screencaps for VIN_+5V and the voltage across the diode D101. Power supply set to 12V (shows ~7mA current draw).

Would adding an RC Snubber fix the issue?