I may be getting greedy, but after researching power supply to use for the output section of F-1, most I am looking at have noise rating in the uV range, and looking at ADP150 the noise is 9uV. To get most out of this I am now thinking I should either feed the 3.3V directly from external PSU or else use LDO which is rated 1 uV or less. The SMT soldering is not something I have experience with, also costly if I need to find hot air rework equipment. Still using the LDO near the board might be worth pursuing. I may try to source already assembled LDO though to minimise how much SMD soldering I need to do.
I guess the question is how much is to be gained by using this setup vs just replacing with the AD150s and feeding an external LPSU are per your original mod. <1uV LDO don't seem to be cheap or many options available.
First a little background about LDO's total output noise: a simple estimation is as followed,
------------------------------------------------------------------------------------------------------
Vout_total = sqrt[ Vout_n^2 + (Vin_n * PSRR)^2 ]
Vout_n: LDO's output noise (9 uVrms @ 10Hz ~ 100kHz; ADP150)
Vin_n: LDO's input source's noise
PSRR: LDO's power supply rejection ratio or ripple rejection (55 ~ 70 dB @ 10Hz ~ 100kHz; ADP150)
(Note: only estimation, in practical integration over the whole bandwidth is required; also coupled noise from other sources must be taken into consideration too)
------------------------------------------------------------------------------------------------------
Take SU-1 as example: it use LM2940 as 5V LDO (54 ~ 72 dB @ 120Hz; output noise = 150 uVrms). Let say its input have 10 mVrms rippe/noise and assume LM2940's average PSRR ~= 60 dB or = 1/1000.
V5v_total = sqrt[ 150u^2 + (10m / 1000)^2 ] ~= 150.33 uVrms
Now look at 3.3V stage, let's also assume average PSRR ~= 60 dB (1/1000)
V3v3_total = sqrt[ 9u^2 + (150.33u / 1000)^2 ] ~= 9.0013 uVrms ---- 5V from LM2940
V3v3_total = sqrt[ 9u^2 + (0u / 1000)^2 ] = 9 uVrms ---- ideal 5V source
Now you can see an only decent LM2940 (150 uVrms) will result to almost identical last stage noise level as a 0 uVrms ideal source. That's why Singxer choose LM2940 (more cost effective).
So depending on complexity, you can;
(A) Remove only isolation/swiching power supply (Mornsun B0505S) and connect (+5V, Gnd) to any decent (<100 uVrms) external 5V power supply; 1uV LPS won't provide any advantage. This is supposed to give you same level of performance as SU-1.
(B) In addition to (A); also replace LDO for CPLD with an ADP150. Because of both LDOs are ADP150, you can comfortably power the two LDOs with Lithium-ion battery. This is my current configuration.
(C) In addition to (A) and remove both LDOs; use ~1 uVrms LPS to directly provide 3.3V power to the XOs. This is what I am planning to do.
This is a pre-assembled module (at least the smd parts) --- "http://www.diyinhk.com/shop/accessories0/73-08uv-ultralow-noise-dac-power-supply-regulator-33v5v-1ax2.html", but right now it seems out of stock.
Right now I can only find two around 1 uV LDO; LT3042 and ADM7154: LT3042 has reverse input protection which is very useful if powered by battery, also its input can range from 3.8 ~ 20 V. Its PSRR @ 10 Mhz is a little below 60 dB. As for ADM7154, its input range is 3.8 ~ 5.5V, but its PSRR @ 10 Mhz is a little over 60 dB. Also its evaluation kit "http://www.analog.com/en/design-center/evaluation-hardware-and-software/evaluation-boards-kits/eval-adm7154.html#eb-overview" is only 49 USD, much more reasonable than LT3042's 150 USD demo board. So there is option (D) replace LDO for XOs with ADM7154 ev kit and power the ev kit and the other LDO with a decent external LPS.