LM317 low input dips & resistor efficiency

[stev3wood]

hi! My dad's been fiddling with this circuit and I'm trying to help him out. He's playing around with an LM317 voltage regulator and I wanna help him and got a few questions.
First off, we're trying to figure out if the output voltage stays solid even if the input voltage changes, as long as it's within the LM317's 3−40V input range, of course. Can anyone confirm that?

He needs a small circuit that will take a constantly changing input of 0−30V and output a stable 5V. but if the input voltage dips below the LM317's minimum 3V input for tiny bursts (like 10−100 milliseconds), is that a problem?
Also, any genius ideas for a small component or fix that could handle those really low input voltages so the regulator always gets its minimum?

Finally, we're trying to choose resistors for R1 and R2 to get that 5V output. We calculated that R1=5.6kΩ and R2=12kΩ give about 5.15V.
Our local supplier is out of the lower value resistors (R1=3.2kΩ and R2=10kΩ), which would also work.
Do the higher value resistors use more power? We need to keep this thing as efficient as possible, so we're wondering if we should wait for the lower value ones.
Any help would be awesome! Thanks a ton!

 

[General_Ebdahl]

The LM317 holds steady only if the input stays above the dropout margin, usually around 2 to 3 volts above output. If the input dips below that even for a short time, the output will sag. Using bigger resistor values wastes less power but makes the circuit a bit more sensitive to noise. A capacitor on the input can help ride through those quick dips.

 

[Alan]

You might also consider putting a low-dropout pre-regulator, such as a buck converter, before the LM317. This can help by buffering the input and ensuring there's enough headroom. Doing so stabilizes the 5V output, even when there are short voltage drops.

 

[barfle]

For low-voltage dips, you might use a diode and capacitor combination on the input. The diode must be able to handle the peak current of your load, and the capacitor needs to not only handle the maximum boltage it will see, but store enough charge to provide current during the period when the voltage is low.

Fun little circuit design problem!

There are many sources of resistors. Mouser, even ebay and Amazon.

 

[CoolBrad2025]

hi! My dad's been fiddling with this circuit and I'm trying to help him out. He's playing around with an LM317 voltage regulator and I wanna help him and got a few questions.
First off, we're trying to figure out if the output voltage stays solid even if the input voltage changes, as long as it's within the LM317's 3−40V input range, of course. Can anyone confirm that?

He needs a small circuit that will take a constantly changing input of 0−30V and output a stable 5V. but if the input voltage dips below the LM317's minimum 3V input for tiny bursts (like 10−100 milliseconds), is that a problem?
Also, any genius ideas for a small component or fix that could handle those really low input voltages so the regulator always gets its minimum?

Finally, we're trying to choose resistors for R1 and R2 to get that 5V output. We calculated that R1=5.6kΩ and R2=12kΩ give about 5.15V.
Our local supplier is out of the lower value resistors (R1=3.2kΩ and R2=10kΩ), which would also work.
Do the higher value resistors use more power? We need to keep this thing as efficient as possible, so we're wondering if we should wait for the lower value ones.
Any help would be awesome! Thanks a ton!

hi steve, the simplest fix for the dips is an Input Capacitor. A 1000μF or 2200μF capacitor on the input should be big enough to easily smooth out a 100ms dip

 

[lupo_audio]

LM317 is a rather old linear voltage regulator which a stable, adjustable output voltage in from 3 ... 40V. The required voltage drop (VOUT-VIN) have to be larger than 2V! So it's not a Low Drop regulator (LDO).The resistors R1 and R2 will control the output voltage. The voltage drop via R1 is the temperature stable reference bandage voltage of appr. 1.25V. R1 should not exceed 300 Ohms for proper operation. It may even work with higher values for R1 as mentioned, but it´s not guaranteed or recommended (if you use LM317 from different manufacturers). Transient voltage drops down to VIN=0 will not be suppressed by the regulator - you need to an
energy storage (Diode + sufficient large capacitor) at the input.