3
\$\begingroup\$

Further to the SEPIC design I asked about in LTSpice SEPIC simulation current spikes, I have a new problem. The output voltage is OK and the inductor current waveform looks much better but the inductor current has a low frequency ringing component. (Ringing might not be quite the right term?).

ringing current waveform

The blue trace is the inductor current and the red trace is the load current toggling between 100mA and 1 Amp.

The detailed waveform of the inductor current looks OK I think.

enlargement of inductor waveform

I have tried adjusting just about every value (inductors, C4, C6 & R8) without success.

The large ringing at power on is also a worry.

I am not sure how to tune the slope compensation components (C6 R8). What should I be looking at?

schematic

Improve this question
\$\endgroup\$
2
  • \$\begingroup\$ I have used this device, and you can easily get bitten by two things: Frequency foldback during startup and the minimum switch on time. You have the converter running at 1MHz; try reducing that (I know it may change other components, but a quick check by changing Rt should show if odd things are going on). Frequency foldback may well be the cause of the startup 'ringing'. \$\endgroup\$ – Peter Smith Jan 2 '16 at 15:35
  • \$\begingroup\$ @PeterSmith Thanks for the suggestions. I tried dropping the switching freq to 500kHz and it made no difference to the ringing. The frequency foldback doesn't take effect until the device starts switching, which seems to be at about 0.8mS, so doesn't explain the power on ringing. My guess is the power on ringing is C4 charging through the two inductors, so I'm not sure this can be avoided - maybe bigger L and smaller C would help? \$\endgroup\$ – David G Jan 2 '16 at 22:19
1
\$\begingroup\$

After much tweaking of the C4, L1 & L2 values and the slope compensation components, I managed to get waveforms that have much less low frequency component. I have settled on slightly more L (27uH), slightly less C (2u2) and much the same C8 & R6. It would be nice if LT offered a design solution that helped with these calcs!

SEPIC Waveforms after tweaking

Note these waveforms are with a sine output load at 400Hz not the step shown in the original post. Green is OP voltage, Blue inductor current and red load current.

Improve this answer
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.