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.

SN6505AEVM: Instability of LDO output

Part Number: SN6505AEVM
Other Parts Discussed in Thread: SN6505A, SN6505B

Hi,

I have modified a SN6505A EVM the following way:

- replaced SN6505A with SN6505B
- replaced WE 750315240 with a WE 760390015

Input voltage is now 3.3V instead of 5V.

Now I have the parts that are rcommended in the SN6505 datasheet (Table 3) for a 3.3V to 5V converter.
So far everything works fine with the isolated converter.

If I'm doing a load step from 0 to 100mA (47Ohm resistor) at the output of the LDO (TPS76701) I get an istable output. The ripple voltage at the output is as follows:
- frequency: around 165kHz
- amplitude (Vpp): around 93mV

With the capacitors populated on the SN6505A EVM I think all the requested requirements of the TPS76701 are met.

Do I need more capacitance at the LDO output to get a stable output?
If yes, what would be a reasonable value?

Or could it be a problem of the combination of the push-pull converter and the LDO?

Thanks and best regards,
Patrick

  • Patrick,

    I'm sorry you are having trouble with this. The 165kHz frequency of the voltage ripple is suspiciously close to the switching frequency of the 160kHz frequency of the SN6505A. I know you stated you replaced the SN6505A with the SN6505B but can you verify the IC for me? Also please double check the resistor value if you don't mind. Being off by an order of magnitude on accident i.e. 4.7ohms would cause higher loading and therefore the relativity high amount of ripple you are seeing. There should be adequate capacitance on the EVM already and everything else should be playing nice together. Would you be able to provide a scope shot of the ripple as well?

    Respectfully,

    Lucas

  • Lucas,

    I have double checked the loading resistor as well as the Push-Pull driver. Everything is as it should. SN6505B is populated (marked with "650B") and LDO is loaded with a 47 Ohm resistor. (approx. 100mA load current)

    The oscillation only occurs when I have a load transient at the LDO output. Startup with the load connected works fine.

    Attached two scope shots showing the load transient from 0 to 100% with the oscillation starting right after the transient (please see below).

    Switching frequency of the Push-Pull converter is close to 420kHz.

    As the output capacitance of the LDO (10uF) is close to the minimum recommendation from the datasheet, I think that I have to increase the value (?).

    Regards,

    Patrick



    Picture 1:
    CH1: LDO ouput voltage (AC coupled)
    CH2: LDO load current
    CH3: LDO input voltage (DC coupled)

    Picture 2:
    CH1: LDO ouput voltage (DC coupled)
    CH2: LDO load current
    CH3: LDO input voltage (DC coupled)

  • Patrick,

    I really appreciate the scope shots. So I see the transition from no-load to 100mA is very fast (<2µs). Unfortunately this fast of a transient would most likely cause this LDO to destabilize and result in the ripple you are seeing. To test for this theory, try ramping up the from 0 to 100mA several times slower i.e. 100µs. You could also try keeping the same transient load conditions but remove the LDO and connect the load straight to the Vout of the rectifier diodes. If the ripple doesn't present itself in this condition, then the LDO was for sure the issue. If your application demands the fast load transient that you have tested with so far, then adding output capacitance to the LDO would help stabilize the output during the transient. Try adding capacitance in 5 to 10µF increments until stabilization improves.

    Respectfully,

    Lucas

  • Patrick,

    I haven't heard from you in a while. Have you been able to resolve your problem?

    Respectfully,

    Lucas

  • Lucas,

    It is definitely the LDO that caused the instability. Adding an other 10uF capacitor at the LDO output solved the problem.

    Best regards,

    Patrick