0

I am currently testing a DC/DC converter component that it meets datasheet specification. Link to component datasheet: http://www.tracopower.com/products/thn15wi.pdf (THN 15-2411, in: 9-36 V, out: 5 V)

This is the circuit:

schematic

simulate this circuit – Schematic created using CircuitLab

Now the problem: enter image description here

The yellow line on the oscilloscope shows the output. It can be seen that at 1/400k intervals, there are voltage spikes. These are larger than the datasheet specification. They are roughly 300-500 mV. How do we remove these? Is it possible, or is our component defective?

Sheepwall
  • 31
  • 3
  • 1
    How close to the device are your probe points? (both signal and ground need to be very close to each other). This could be an artifact of your measurement. – Peter Smith Jun 21 '17 at 15:16
  • 1
    Looks an awful lot like ground lead bounce to me, indeed. – Asmyldof Jun 21 '17 at 15:22
  • 2
    Show scheatic. Show layout. Show measurement setup. Also watch: https://m.youtube.com/watch?v=Edel3eduRj4 – winny Jun 21 '17 at 15:37
  • 1
    Related: https://electronics.stackexchange.com/questions/102617/smps-output-ringing/102667#102667 https://electronics.stackexchange.com/questions/199244/ripple-and-ringing-from-cheap-chinese-boost-converter/199261#199261 – Adam Lawrence Jun 21 '17 at 15:39
  • 1
    1. Never measure ripple without BWL turned on - your Agilent 5/6000 scope shot shows it off. 2. Always measure with a short-ground probe or clip with local decoupling (small ceramic + small electrolytic) at the measurement points to eliminate noise pickup - what remains is "real" noise. Also see http://www.techni-tool.com/site/ARTICLE_LIBRARY/Agilent%20-%20Tips%20and%20Techniques%20for%20Making%20Power%20Supply%20Noise%20Measurements%20with%20an%20Oscilloscope.pdf – Adam Lawrence Jun 21 '17 at 15:44
  • 1
    It seems weird to me that your digital scope has no screenshot feature and you have to take a pic of the screen... Anyway, show a pic of your layout, with probes visible in measurement position. Also, a sanity check is to probe GND. If you have the same ripple on the output and on GND, question your measurement technique... – bobflux Jun 21 '17 at 16:39
  • I bet it is some breadboard with 4" leads and a cheap aluminum electrolytic cap in between. You can't evaluate a 400kHz 3-A switcher with 28 V input on a breadboard. – Ale..chenski Jun 21 '17 at 18:40

0 Answers0