Author Topic: Discrete ADC LDO placement on ADC  (Read 415 times)

0 Members and 1 Guest are viewing this topic.

Offline electronxTopic starter

  • Regular Contributor
  • *
  • Posts: 146
  • Country: 00
Discrete ADC LDO placement on ADC
« on: May 13, 2024, 12:21:47 pm »
I am making a design that includes discrete ADC. I use an LDO to feed the ADC. I also feed the LDO with SMPS. Why did I use LDO before ADC? Because I want to suppress the ripple. That's why I used an LDO that will suppress the switching frequency of the SMPS before the ADC, as in the analog devices and TI documentation. (Even though the ADC has a high PSRR, I am measuring at the uVolt level, so I do not want to see any switching noise around the ADC).

My question here is, is it better to move LDO closer to ADC or keep it closer to SMPS? If we think of it with the logic of finding and eliminating the noise at its source, having the ldo close to the SMPS seems like a better option. So, as I showed in the image, is it better to move the ldo closer to the a or b direction? Any ideas?



in god we trust
 

Online moffy

  • Super Contributor
  • ***
  • Posts: 2105
  • Country: au
Re: Discrete ADC LDO placement on ADC
« Reply #1 on: May 13, 2024, 12:56:59 pm »
I vote for nearer the ADC as the ground will be closer to the ADC ground, also keep the SMPS switching currents near itself.
 

Offline mawyatt

  • Super Contributor
  • ***
  • Posts: 3895
  • Country: us
Re: Discrete ADC LDO placement on ADC
« Reply #2 on: May 13, 2024, 02:54:47 pm »
Agree.

The best location for a regulator is at the load, or point of service generally speaking. This goes for on-chip, PCB, Chassis and Systems.

Best
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 

Offline MasterT

  • Frequent Contributor
  • **
  • Posts: 826
  • Country: ca
Re: Discrete ADC LDO placement on ADC
« Reply #3 on: May 13, 2024, 05:26:07 pm »
I see LDO & SMPS, likely ~1V for drop-off margin? 
My recent discovery that LDO makers cheat in datasheet regarding PSRR and drop-off voltage vs load current. Two parameters "mutual exclusion", means if one is important - disregard another one. To have PSRR >60dB great majority LDO regulators demands 3-5 V  Vin - Vout difference.  No one provides PSRR vs Vin-Vout in data sheets, here some results I get:

// TO-220  Out=3.3V & Vp-p=2V, F-1kHz

                  Diff=Vin-Vout
1. LD1117           -  3.7V
2. LD33             -  3.7V
3. LD1084           -  2.4V
4. LM317T(ne LDO)   -  3.0V
5. LM317T( LDO)      -  2.7V
6. LM317AHVT      -  2.7V
7. LM2941+         -  4.5V ** Out=5V min. Vp-p=1V, Vin=9.5V
7. LM2991-         -  1.2V *"bump" 10kHz -48dB (2kHZ -60dB)



               V-diff  PSRR
            Volts.   dB-1 kHz
   AP7381      1.5      -60
   LM2931          1.7     -54
   LP2950-3,3   1.7      -70
   L78L33       2.7      -80  * not LDO
   MCP1702       1.7      -18
   S-812C-3.5  1.3     -19     No comments. 1uF+1uF
   S-817B-1.8  0.8     -29     1uF+1uF

//  SOIC, test:  1uF+10uF, 15 mA, 1 kHz, Vout = 2.5 В
   NCV47701    2.5     -60
   NCV47551    1.9     -51     NR = 0.00uF.
   NCV47551    1.7     -70       NR = 0.22uF.
   LM2931A     0.28    -44
   LM2931A     4.0     >60
   AP7361A     0.1     >60       5 star *****! 100мВ guaranteed PSSR.

    NCP731      0.6     >85
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf