Yet another ultra low noise high PSRR LDO - LT3042

Yes.
As on the pic here..
 

Attachments

  • Screenshot_20211011_140315_com.android.gallery3d.jpg
    Screenshot_20211011_140315_com.android.gallery3d.jpg
    274.1 KB · Views: 276
Thank you so much. You help was so welcome, now i will use Cset=22u+150k, no Rlim, and i am thinking of using the PGFB divider as the minimum load. Its the best way i think to use less resistors, good idea ?


Because i dont think a single opamp is enough to be in the best current zone for the 3042.
 
Last edited:
Jan,

Visibly the LT3042/LT3045 are so fast that the equivalent output inductance do not go in resonance even with low ESR output capacitance.
See scan of the output impedance; the important part is the minimum load current (for the velocity)
here it was 60mA.

Trace 1 is the actual active regulator ootput;
Trace 2 (less bright) is the regulator 'off', so principally one sees the output bypass capacitor..
 

Attachments

  • IMG_20200331_040008.jpg
    IMG_20200331_040008.jpg
    711.1 KB · Views: 170
Last edited:
Perfect: the more the better.. :)
100/200mA..but you do not have headroom left..

If remember well (I had done these measurements in 2020..lockdown..) - then smthing like 30/35 mA could be enough for getting into a usable range. (speed, output impedance)
I keep 50mA as a rule of thumb..
 
AX tech editor
Joined 2002
Paid Member
Jan,

Visibly the LT3042/LT3045 are so fast that the equivalent output inductance do not go in resonance even with low ESR output capacitance.
See scan of the output impedance; the important part is the minimum load current (for the velocity)
here it was 60mA.

Trace 1 is the actual active regulator output;
Trace 2 (less bright) is the regulator 'off', so principally one sees the output bypass capacitor..

Yes, I see, these regs are something else.
Very high performance and ease of use.

Jan
 
One more thing: the copper area beneath the chip must be large enough for the dissipation. LT30xx are quite robust and the internal power and thermal limiting circuits work well. When these kick in the performance starts to suffer. This performance degradation may however be missed depending on the application.
 
What I refer to "SilentSwitcher" is dual boost/inverter (e.g. LT3471 or LT8582) that generates dual supply from single supply (e.g. 5V in to +-15V out). These output voltages are then regulated with low noise regulator. Jan's original SilentSwitcher uses TPS7Axx regulators. My take on this uses LT3045/LT3094. It has lower lf noise but otherwise not that much better.