Guys, did you proceed with this issue? I have exactly the same problem. Thanks to this thread I found issue with ground plane connection Cin - IC GND went through vias and it was not enough low impedance connection. When soldered extra wire then Vout increased but not to supposed level.
I still could see irregular switching cycles at 260kHz, correct-spike-spike-correct. Cycles were dependand on resistive load 5r6-1r4. The only solution to get it working is hack with capacitor across diode. Even temporary capacitor to start function is enough when I remove it later then continues to delivery power. When inspecting V_switch then I see different wave forms after every power cycle. Sometimes nice regular equal rectangles, sometimes rectangles with alternating width when likely area of active period is equal to area in regular care in particular time. But the output voltage is stable and at supposed level 5V1. I see high-freq oscillation 78MHz in output voltage with high amplitude about 1-1.8V after switch. What LC circuit has influence on it?
I suspect "cycling issue" might be related to sensing max. current when some spike is detected as overcurrent and IC switches off FET. Feedback voltage seems stable - on the other side 78MHz oscillations should be there too but I cannot see anything dramatic. Also in block scheme is a RC which likely integrates feedback to remove spikes. I do not see more option how to tune it. Softstart is likely acting after power cycle only.

Mod: Link to old thread.
https://forum.allaboutcircuits.com/...ble-to-reach-full-voltage.159369/post-1387494

 

Guys, did you proceed with this issue? I have exactly the same problem. Thanks to this thread I found issue with ground plane connection Cin - IC GND went through vias and it was not enough low impedance connection. When soldered extra wire then Vout increased but not to supposed level.
I still could see irregular switching cycles at 260kHz, correct-spike-spike-correct. Cycles were dependand on resistive load 5r6-1r4. The only solution to get it working is hack with capacitor across diode. Even temporary capacitor to start function is enough when I remove it later then continues to delivery power. When inspecting V_switch then I see different wave forms after every power cycle. Sometimes nice regular equal rectangles, sometimes rectangles with alternating width when likely area of active period is equal to area in regular care in particular time. But the output voltage is stable and at supposed level 5V1. I see high-freq oscillation 78MHz in output voltage with high amplitude about 1-1.8V after switch. What LC circuit has influence on it?
I suspect "cycling issue" might be related to sensing max. current when some spike is detected as overcurrent and IC switches off FET. Feedback voltage seems stable - on the other side 78MHz oscillations should be there too but I cannot see anything dramatic. Also in block scheme is a RC which likely integrates feedback to remove spikes. I do not see more option how to tune it. Softstart is likely acting after power cycle only.

Mod: Link to old thread.
https://forum.allaboutcircuits.com/...ble-to-reach-full-voltage.159369/post-1387494

What device are you using? What is your input voltage? What output voltage are you observing? What is your load? Do you have enough input decoupling close to the IC?

 

It was post to 5 years old thread so new thread has been created and context discontinued. Old issue but the same symptoms.
It was thoroughly discussed there:

Excerpt and update:

setup
LM2679, Vin=24V, Vout=5V1, Ioutmax=4A, soldered on PCB, circuit from TI workbench, Isense 6k19 (I peak max. 37.125/6.19=6A), resistive load, input power from commecial 24V adapter, no problem with Vin - is stable, coil https://www.we-online.com/components/products/datasheet/7447709150.pdf
problems

• working under low load normally, resp. output voltage as expected
• with load output voltage dropping down
• subissue: 78MHz oscillation after switch

observations

• Vsw there are irregular square wave widths in waveforms at 260kHz

suspicion about reason

• Imax sensing issue
• unprobably startup issue
• unprobably Vfeedback issue

fixes

• better ground plane among Cin, Cout, D, IC
• 100pF capacitor bridging diode - hack from original thread
• decoupling (C205) could be closer to pin #2, now is nearby Cin (C204), will be fixed in next gerber. Manually held 100nF capacitor just between #2 & #4 of IC and not visible change in waveforms

question

• any progress with finding root cause of problem (as capacitor is hack) ?
• what is supposed LC circuit causing resonance at 78MHz

pcb fix


without capacitor, not able to deliver supposed Vout blue, 3.64V@1R4ohm, yellow Vsw


with capacitor, Vout 5.040V@1R4ohm, irregular but wider ON phase, ramdomly even equal rectangles happen in particular power cycle

resonance detail


schema

 

This is TI's recommended layout. Notice how close the supply decoupling capacitor (Cin) is to the device:
I've used the LMR38010, and it really does make a difference.

 

Yes, it is layout from datasheet (BTW they might provide color layout and not grey picture where silk color is mixed with gaps), but I'm limited by Z space for electrolyte caps because there is stacked RPI5 with fan and I also require Vin and Vout on right side.

Meanwhile I tested previous PCB design v1.1 and it works better. The main difference is likely feedback trace which does not go nearby coil. I can achieve regular waveform easily. Seems even after Cin-#4 wire fix without capacitor hack. Sometimes I have nice regular waveform, sometimes not. But it can also change probe capacitance so I don't know what is bare unprobed state.

Comparison of 2 layouts (left is new)


waveforms from v1.1:

perfect waveform


unbalanced but Vout OK


before wire fix soldered, ugly spikes but Vout was not too bad as there were 2 wide forms and 1 spike

 

When I reverse engineer proposed layout I see there is undocumented stuff (cyan).

1. What are RdCd parts between SW and GND, looks as RC circuit discussed in original thread, i.e. our hack capacitor in series with resistor
2. How is connected Vout to the top of divider for feedback. Maybe it is not ADJ version ...

Aha, all answers are in evaluate board datasheet https://mm.digikey.com/Volume0/opasdata/d220001/medias/docus/2978/LM267x Eval Board.pdf




The evaluation board was designed primarily for circuit implementation using all surface-mount components. The small series "trace inductance", particularly from the Switch Output pin, can create a high frequency (10's of MHz) ringing signal at the switch output. If problematic, this ringing can be reduced or eliminated by the use of a series RC damper or snubber network from the switch output to ground. The addition of these components is made at the locations labeled CD and RD. Values of 0.01 µF and 10Ω are good starting values that may need to be varied depending on the magnitude of parasitic factors in a given design. In an actual end application, these components are normally not required if proper care to minimize trace lengths is taken in the PCB design.

Even I have shielded inductor:
If open core inductors are used, special care must be taken as to the location and positioning of this type of inductor. Allowing the inductor flux to intersect sensitive feedback, IC groundpath and COUT wiring can cause problems.

When using the adjustable version, special care must be taken as to the location of the feedback resistors and associated wiring. Physically locate both resistors near the IC, and route the wiring away from the inductor, especially an open core type of inductor.

So seems main my issue in v1.3 is feedback trace and divider located wrongly on the other end of trace ! Plus I will add RC across diode

Thanks you pointing me to proposed design engineering which I originally considered as grey puzzle![/QUOTE]