search

sinara-hw / Urukul

4 channel 1GS/s DDS (AD9910 or AD9912 variant)
14 stars 8 forks source link

Urukul LDO PSRR #28

Closed hartytp closed 5 years ago

hartytp commented 6 years ago

@gkasprow Looking at figure 11 of the TPS74901 datasheet, the PSRR of regulators IC7, IC11 and IC18 is very low with a dropout voltage of 0.2V. Can we change the outputs of IC15 to give an extra couple of hundred mV headroom on these LDOs?

Edit: this is probably also a problem for clocker...

jordens commented 6 years ago

What are you after? The 1 MHz of the SMPS were already down 85 dBc and comparable to the DDS spurs. And at 1 MHz the LDO PSRR doesn't change much with dropout. Give it another 100 mV. But 300 mV would add another 10% in power consumption.

hartytp commented 6 years ago

More worried about general broadband crud coming through from the SMPS.

We're planning to do a careful AM/PM noise measurement, hopefully tomorrow. From that, I'll know if this is actually needed...

jordens commented 6 years ago

Except for the 1khz Spurs you'll measure the phase noise that I measured already.

hartytp commented 6 years ago

With the PLL on, this won't help. Not sure about the noise when the DDS is clocked directly from a low noise 1GHz oscillator.

In any case, let's not do this until we have evidence that it will help.