Rectifier ripple on DC-DC buck converters

Question

I'm using the LMZ36002 buck converter module, which is a 1 MHz 4.5-60 V 2 A buck module. This means there is already an integrated inductor in the chip package. This device will accept any voltage between ~6V and 60V and regulates it down to my desired 5 V, 1.5 A board supply.

Normally, this device is run on 24 V DC from line power supply. Now I would like to add AC functionality by rectifying and smoothing the input.

My calculations are as following: (efficiency corrections are ignored)

• 5 V @ 2 A = 10 W
• 24 VAC * 1.414 = 34 V
• 10 W @ 34 V = 300 mA

I've simulated a circuit with diodes and one 470 uF capacitor in LTspice.

The load is a 70 ohm resistor. The ripple voltage on the rectified output is quite large, around 6-8V, but well within the buck converter's input range. Since space is an issue in this project, low capacitance like 470 uF would be preferred.

• Will this large ripple voltage cause any problems for the DC-DC converter?
• Could there be any serious problems, aside from degraded efficiency?

Answer 1

As long as the LMZ36002 operates within the recommended input voltage limits, it should generate the required DC output voltage without a problem. As you correctly noted, the efficiency is dependent on the input voltage. A point you will need to investigate is line rejection. Since there is a significant ripple component on the input voltage, a small component of may propagate to the output. It all depends on the modulation strategy of the DC-DC converter and also the bandwidth of the voltage control loop. A quick test should reveal whether the converter has good line rejection or not. My guess would be that it has good line rejection since the switching frequency is so high and there are no external components used in the control loop. This would tend to indicate that the bandwidth of the internal controller is high and it can respond rapidly to changes in the input voltage. You'll probably only notice a few mV of ripple.

Answer 2

There is no specification of PSRR nor graphs of transient response to line changes. They only give one performance spec for the output:

Based on the lack of datasheet coverage I would either test it thoroughly and/or simulate it, OR pick a different part.

TI usually does a great job with their parts, but I've used a different small integrated regulator from them and it doesn't have any line regulation specs and it doesn't do well with line transient regulation.