• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 1208 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Playing with LM317/337 - few concern in a new design

Mateusz Kaluzny17
Expert 1065 points
Other Parts Discussed in Thread: LM317, TL431, LM337

Hello dear Ti community.
I was using LM317 for a quite long time in my cheaper circuits.
But then I was wondering about something better and LM317 was leaved behind due to TPS7A regulators and some other stuff from Linear Tech.

But after all there is much more to do with LM317 to improve it performance in term of PSRR, noise or output impedance.

A I've go some question related with those mostly.

1) how about CADJ? Should it be special kind? I mean tantalum (solid, polymer) electrolytic (ESR range?) or classic ceramic (X7R), or maybe something else?
There is a plenty of non popular capacitor called as a film, builded around plastic or acrylic that could be used there.
Even big film capacitors (there is a possibility to buy 10uF film cap with 5mm raster).
I know this is kinda old regulator, and datasheet is mostly copy-paste of old paperwork from that time.

2) how about replacing R2 with a zener diode or even shunt regulator (TL431 for that matter) it should improve noise performance, because dynamic impedance of a zener or shunt regulator is MUCH lower than a R2 itself.
Or maybe there is a better way to improve this aspect of performance? How about TL431 will it work well, or maybe is should pick something else?

3) how about lowering output impedance to avoid any resonance that might be pronounced by a capator itself?
LM317 output impedance is kinda high and it will rise quickly just before 1Khz.

In reagard to this document, method described there should help me with this aspect, but what do you think?

Another important thing here is what of capacitors should be used for best performance?

4) question about "shutdown" option, as I remember it was descibed in a datasheet, I assume that when 0 level will be pushed out then regulator itself will output voltage equal to it's reference? 
In this case it will be 1.25V ?

5 is there exist any kind of circuit that could be used between LM317 and LM337 that will be mutual shutdown circuit, that will halt regulators if one rail collapse?
I saw something like than in a Douglas Self book, but he was using both positive regulators with separate transormator windings.

Just for your reference I will post a picture that will show few things in regard to my questions.


  • 0
    TI__Guru 55940 points
    Hi Mateusz,

    I apologize for the delay in this response.

    As you noted, LM317 is over 20 years old and while it was competitive at the time it released, our newer LDOs have better performance. While application notes are available on many of the topics that you mention, they have been written for our newer portfolio of linear regulators and LDOs.

    The optional Cadj does not need to be a specific kind of capacitor; however, a X7R capacitor will have the advantage of having relatively good capacitive stability across temperature. The datashet figures that you point to show how this additional filtering improves PSRR.

    It would not be recommended to replace R2 with a diode or shunt regulator. Current must flow through these resistors in order to set Vadj. The internal amp drives the pass element until Vadj is equal to the internal reference voltage. Adding diodes or other active circuitry to this feedback loop will impact the loop response and stability of the linear regulator.

    If ADJ is tied to OUT, then Vout will be equal to the internal reference of 1.25 V.

    Neither LM317 nor LM337 have an enable pin built into the device. These regulators are designed to output a voltage whenever an input voltage is present. They will operate in dropout until the input voltage is high enough to regulate the output at their set voltage. The easier location to add a disable would be at the upstream devices supplying the input voltages to these regulators.

    Very Respectfully,
    Ryan