LM27762: PSRR value @2Mhz of Positive LDO

Hi Gerald,

Unfortunately we only have data up to 100KHz. The positive LDO doesn't have enough bandwidth to actively improve the PSRR at high frequencies. At 2MHz, the only attenuation an input noise signal gets is from the RC filter formed by the resistance of the LDO pass transistor and the output capacitor. Does the customer need a certain minimum level of PSRR at 2MHz?

Best regards,

Varun

Hi Varun,

Thanks for the initial reply. Our customer need a bias supply for op-amps, -/+3.3V, 50mA. Behind LM27762, there is a 4V out, 2MHz, Buck converter. So they need to know how much of buck converter's ripple will be seen at the power pins of op-amps. May we ask what do you think, better using LM27761+TLV740P, which both parts PSRR defined @2MHz?

Also there is no information about LM27762 positive LDO's series resistance. Is it ok to calculate it from V-I graph which is 0.48 ohms ?

Kind regards,
Gerald

Hi Gerald,

Is the buck converter supplying the 4V rail running in PFM mode? That would mean it has some low frequency content in addition to the switching noise.

Adding a 2nd LC at the output of this buck converter can help attenuate the high frequency switching noise. I'm assuming this buck switches at 2MHz? You need to design the cut off frequency of this 2nd LC outside the bandwidth of the Buck converter. Assuming the bandwidth is below 100KHz and you design a 2nd LC with a cut off around 200KHz, you can get around 40dB of PSRR at 2MHz.

If it is important to filter the low frequency noise from the PFM switching, an LDO can help here. The LDO also helps at high frequency (2MHz) due the passive RC filtering. The amount of filtering really depends on the operating conditions as the internal LDO resistance depends on VIN, VOUT and load current. The larger the internal series resistance, the better the LDO PSRR is at high frequency. This is because you move the cut off frequency of the RC to lower values . Larger VIN/VOUT ratios and smaller load currents result in higher series resistance and better filtering. The internal series resistance can be calculated as (VIN-VOUT)/IOUT.

Since it's passive RC filtering at 2MHz for most LDOs (unless you find a high bandwidth LDO), I don't expect to see any difference the PSRR of different LDOs operating at the same VIN, VOUT and IOUT.

Coming back to this application, I would recommend a 2nd LC at the output of the buck for ~ 40dB PSRR at 2MHz. If you need to attenuate the low frequency PFM content or even more attenuation at 2MHz, an LDO can be used.

Best regards,

Varun

Hi Varun,

Thanks for all the effort and information you provided. In our customer project, they need -3.3V, so the buck converter behind should be above 3.3V. They decided a value around 3.6V to 4V according to the drop outs etc. So they already need a clean -3.3V, and two +3.3V's one for analog and other for mcu. They also need a clean +30V for HV opamps. Here they think to use TPS7A4101. At this time he want to talk on part number basis. The buck converter can be TPS560430 in FPWM mode @2MHz. They want to make a decision between these two: 1- Using a single LM27762 for opamps and TLV74033 for mcu. 2- Using LM27761 plus one TLV74033 for analog, one TLV74033 for mcu. Analog requires 20mA. MCU requires 150mA

In this case ( he mean they already need to use 3xLDO's in total) do you think do they still need additional filters ?

Regarding HV op-amps regulator, behind TPS7A4101 there is UCC256404 based LLC converter operating at 100kHz
If anything is not clear please don't hesitate to ask.

Kind regards,
Gerald