0
\$\begingroup\$

What could be the reason of such spikes in inductor current at each Switch ON-OFF. ??

This spikes are also getting affected in load current.

Please help me to reduce / remove this spikes.

( waveform of inductor current with spike are shown in image file ) enter image description here

\$\endgroup\$
6
  • 2
    \$\begingroup\$ How are you measuring the inductor current? Those spikes look like noise coupled into the measurement from the switching of the transistor. \$\endgroup\$
    – Dave Tweed
    Commented Jul 19, 2014 at 14:46
  • \$\begingroup\$ The oscilloscope picture has not discernible units of amplitude or time and ditto what dave says above. \$\endgroup\$
    – Andy aka
    Commented Jul 19, 2014 at 14:52
  • 1
    \$\begingroup\$ Try adding a RC network as a low-pass filter at the gate of Q1 to smooth out the control action. \$\endgroup\$
    – Evan
    Commented Jul 19, 2014 at 20:01
  • \$\begingroup\$ I used CT - (Current Transformer) to measure this Current. I am using IGBT as switch @ 8 khz with 1.1mH Ferrite Core Inductor. My load ratting is 160v,7A. Inductor current ripple is 8.5 amps peak-peak. The Noise frequency is ~3Mhz. \$\endgroup\$
    – vasant
    Commented Jul 21, 2014 at 14:36
  • \$\begingroup\$ My load is UV-Tube which is sensitive to current. Due to this resonance spikes my tube is getting flickered. please help me. \$\endgroup\$
    – vasant
    Commented Jul 21, 2014 at 14:42

3 Answers 3

Reset to default
1
\$\begingroup\$

At which end of the inductor are you measuring the current?

If you have put the CT at the Q1 end of the inductor there is a possibility you will couple the fast transitions into the measurement signal.

To minimize this put the CT on the Iout end of the inductor.

The parasitic capacitance of the inductor could also be causing the problem. To ensure this does not pass through to the load it is vital that the output capacitor has an extremely low parasitic inductance and that the circuit layout is well organized.

To deal with the 8.5A ripple the output capacitor should probably be composed of multiple ceramic capacitors in parallel to get the lowest series inductance.

kevin

\$\endgroup\$
1
  • \$\begingroup\$ :Is problem due to "parasitic capacitance of the inductor" OR "parasitic inductance of capacitor" Please clear me to resolve problem.. \$\endgroup\$
    – vasant
    Commented Jul 23, 2014 at 5:18
4
\$\begingroup\$

Unless you have an (expensive) current probe, it's pretty hard to make accurate measurements of inductor current. I'll assume you have a small resistor in series with the Vout end of the inductor. It's also not apparent what the switching frequency or ringing frequency is.

If it's real, it could be ringing of the inductor at its self-resonant frequency (SRF).

One way to deal with that is to use an inductor with a higher SRF (at least 10-20x the switching frequency. Another is to slow down the switching edges so that the energy content at the inductor SRF is low. It reduces efficiency, but it can be a good trade-off to reduce EMI. A series gate resistor is commonly used.

Even if the noise in your output is due to other reasons, slowing the switching transitions should reduce it.

\$\endgroup\$
4
  • \$\begingroup\$ Hello @Spehro Pefhany sir, I used CT - (Current Transformer) to measure this Current. I am using IGBT as switch @ 8 khz with 1.1mH Ferrite Core Inductor. My load ratting is 160v,7A. Inductor current ripple is 8.5 amps peak-peak. The Noise frequency is ~3Mhz. (SFR) 1. my SFR is already @ too high freq !!! My load is UV-Tube which is sensitive to current. Due to this spikes my tube is getting flickered. 2. I used 47 Ohm gate register with gate pulses +15 & -5. please help me. \$\endgroup\$
    – vasant
    Commented Jul 21, 2014 at 14:53
  • \$\begingroup\$ @Spehro Pefhany When you say real, do you mean if it is truly part of the circuit or this is just being observed because its being measured? \$\endgroup\$
    – Funkyguy
    Commented Jul 22, 2014 at 13:21
  • \$\begingroup\$ Maybe you need to add an LC filter or post-regulator after the capacitor. What does the tube voltage and current look like? (numerically and waveform)? And what is it that you actually need?? Getting really low noise from a SMPS is usually not possible without some kind of post-filtering or post-regulation. 3MHz should be easy to filter out. \$\endgroup\$
    – Spehro Pefhany
    Commented Jul 22, 2014 at 14:19
  • \$\begingroup\$ @Spehro Pefhany sir, waiting for your reply..!!! \$\endgroup\$
    – vasant
    Commented Jul 24, 2014 at 3:30
0
\$\begingroup\$

Check the voltage waveform at the VIN side of the inductor. If VIN is not sufficiently bypassed, it could ring when the converter switches. This would show up on the IL waveform as well.

\$\endgroup\$
2
  • \$\begingroup\$ what do you mean by "VIN is not sufficiently bypassed" ??? \$\endgroup\$
    – vasant
    Commented Jul 23, 2014 at 5:21
  • 1
    \$\begingroup\$ The schematic diagram in the original question shows a bypass capacitor across the power input terminals labeled "VIN". If there is not enough capacitance there, or wires to the capacitor(s) there are so long that the parasitic inductance makes them ineffective -- in other words, if VIN is not sufficiently bypassed -- then the circuit could ring when the converter switches. \$\endgroup\$
    – davidcary
    Commented Apr 30, 2015 at 4:19

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