• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 61 subscribers
  • Views 420 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM5175: FETs + RC nets to prevent false OVP really needed?

Thomas Nilsson
Prodigy 195 points
Part Number: LM5175
Other Parts Discussed in Thread: LM3489

Hi

In 's article "Adjusting VOUT in USB Type-C™ and wireless charging applications, part 1" (link below),

there is an example presented to use switched resistor branches to program additional VOUT levels (Figure A)

However, due to the risk of triggering false OVP events when switching, another design is proposed, utilizing a FET and an RC-net (Figure B)

So, my question is, instead of incorporating three extra FET's and three extra RC nets, wouldn't it be possible to add a Capacitor like in Figure C? If not, why not? If so, how is a suitable value calculated?

And thus being able to use open drain outputs on an MCU to sink the resistor required for the wanted VOUT to GND.

Thanks, Thomas

Figure A:

Figure B:

Figure C:

https://e2e.ti.com/blogs_/b/powerhouse/archive/2020/02/10/adjusting-vout-in-usb-type-c-and-wireless-charging-applications-part-1

  • 0
    TI__Guru 67150 points

    Hello Thomas,

    Would you please again attach the 3 figures? They did not get attached the first time.

    If you have an MCU available to modulate the FB pin and the output of the GPIO is able to source and sink current, then you could use the method used in the app note: 

    There the FB voltage is modulated by a PWM signal coming out of the used MSP430. But dependent on the accuracy you want to achieve a slow feedback loop through the microcontroller to adjust the PWM duty cycle might be necessary.

  • 0 in reply to Brigitte
    Prodigy 195 points

    Hi

    Ok. I'll see if I can re-post.

    However, I believe the design used on the TPS65987EVM board, which has the LM3489 buck controller, uses the method I suggested above. On that board there's a 1000pF cap C58. I made a quick test with the LM5175 and it seemed ok so far.

  • 0 in reply to Thomas Nilsson
    TI__Guru 67150 points

    Hello Thomas,

    Ah, ok, you meant to just use one RC filter and then use several switches without dedicated RC filters. Yes, this is possible as well. Please be aware that the transient response of the converter is in this case influenced by C58. If there is a load transient on the output, the voltage loop of the converter is relatively slow due to the capacitor from FB to GND.

    With the solution Youhao proposed in his article, there is no capacitor added to the FB loop (only the parasitic capacitance of the transistor), so the influence to the load transient is a lot lower.

    Another possibility would be a third resistor on the FB node and then a voltage source connected to it (the voltage source could be a GPIO generating a PWM signal followed by a RC filter, or a DAC output).

  • 0 in reply to Brigitte
    Prodigy 195 points

    Yepp, that was my intention :)

    It was in the pics that were missing ;)


    Thanks!