0
\$\begingroup\$

I've recently got my PCB developed for a power supply unit using the LM5021-1. After testing i've concluded there's problems in the circuit for the Vcc pin. I've used the following circuit for this, which is basically a copy from an example inside the LM5021 datasheet (exact copy apart from having to replace some old components):

Used circuit on PCB from KiCad schematic

Now my test results so far have looked like this (measured on Vin_stab label):

Results measured with oscilloscope

This keeps repeating itself. As seen here the lowest voltage reached is around 7V and the highest voltage is around 20V. This fits according to the datasheet: The 20V is when the LM5021 starts while the 7V is when it shuts down. The circuit is supposed to go above 20V and after the AUX winding starts powering up it will take over and stay around 14V. I simulated this circuit in LTSpice aswell (slightly different components) and got the following result:

LTSpice result

I've tried adding a capacitor in parallel (basically going from 1uF to 11uF) to stay at 20V for a longer period but this didn't seem to resolve the problem so far. I've also tried without a mosFet on the output of the LM5021-1, which should result in the AUX winding not powering up and the circuit never lowering to 14V as intended but instead staying above 20V. This ended with the same result of turning on and off.

I've seen another solution using only high value resistors (1M and above) in series from 325V, a diode from the AUX winding and a capacitor. But i am wary of using this since i did not get the simulation to work this way.

My question is why the PCB result is different to the simulated result and how to resolve this?

Link to datasheet: https://www.ti.com/document-viewer/LM5021/datasheet. Schematic is posted in paragraph 8.2

Also here is the full schematic: Transformer + LM5021 Vaux and Vaux_O are connected by jumper. (Same for other "_O" labels)

\$\endgroup\$
5
  • \$\begingroup\$ So what is Vaux actually doing? It seems as though it isn't functioning correctly. \$\endgroup\$
    – Dave Tweed
    Commented Sep 15, 2021 at 11:17
  • \$\begingroup\$ A full schematic of your flyback converter is needed and, try and use proper symbols for BJTs. Have you tried adding a load to the DC output? Add a data sheet link into your question too and state which page you got the circuit. \$\endgroup\$
    – Andy aka
    Commented Sep 15, 2021 at 11:19
  • \$\begingroup\$ @Andyaka I've added the information you asked for, i've not tried adding a load to the flyback converter output. Maybe some additional information: I destroyed the mosFet (Q7 in circuit) when i tried adding more capacitors to "Vin_stab" (not sure why this happened though). \$\endgroup\$
    – Parimo
    Commented Sep 15, 2021 at 12:14
  • \$\begingroup\$ Your device is shutting down on output over-voltage because you haven't applied a load. That's what to me is causing your sawtooth waveform. Try adding a 10% load. \$\endgroup\$
    – Andy aka
    Commented Sep 15, 2021 at 12:23
  • \$\begingroup\$ Thanks for the help! I'm gonna give that a go as soon as i get my last few components, should know more soon. \$\endgroup\$
    – Parimo
    Commented Sep 15, 2021 at 15:48

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.