LDO peak-to-peak voltage 280mV

[stafil]

I have an LD1117V33 connected to a 12V powersupply. When I plug the output to my osciloscope I see about 280mv peak-to-peak voltage, sometime more.

Is it normal? Is there anything I can do to improve it?

My circuit is pretty simple. 12V - LDO - Output with 2 capacitors in parallel in input and output. 100nF input and 10uF output.

[mvs]

When I plug the output to my osciloscope I see about 280mv peak-to-peak voltage, sometime more.

Is it normal? Is there anything I can do to improve it?

Timebase 2µs/div? These noise peaks are quite fast... LD1117 has no chance to cope with them.
It is normal, if you use switching PSU.

Ferrite beads and additional ceramic caps may help you. The only question is, if you really need clean supply.

[stafil]

I've seen the noise go up to 500 pk to pk. I am afraid it will be out of the spec for my microcontroller.

[wraper]

This probably does not even have anything to do with LDO but how you are doing measurement. Say long probe ground lead and noise induced from power source.

and 10uF output

10uF of what type? If it's MLCC, LDO may oscillate.

[stafil]

Yeah MLCC

[Zero999]

This probably does not even have anything to do with LDO but how you are doing measurement. Say long probe ground lead and noise induced from power source.

and 10uF output

10uF of what type? If it's MLCC, LDO may oscillate.

Going by the data sheet, it should be stable, with a low ESR capacitor. It's not a very low drop-out regulator, because it has an NPN output device. I think the ripple is coming from whatever is connected to the input.
https://www.mouser.co.uk/datasheet/2/389/ld1117-974075.pdf

I've seen the noise go up to 500 pk to pk. I am afraid it will be out of the spec for my microcontroller.

I doubt your microcontroller will be damaged, but it could cause problems with analogue circuitry.

The ripple is probably coming from your 12V supply, rather than the linear regulator. You could try an RC filter in the output of the 12V supply. A 1000µF capacitor and 4R7 resistor will have a cut-off frequency of just under 34Hz and will greatly attenuate the ripple which is in the 1MHz region.

EDIT:
Make sure the resistor has a high enough power rating. P = I²R, so for the full 0.8A, the regulator is capable off I = 0.8²*4.7 = 3W, but use a much higher rating. I'd recommend  at least 8W, as the regulator might not current limit until 1.3A. You could also use three 1R8 3W resistors in series, with a 470µF capacitor to 0V, after each, to spread out the power dissipation and provide a steeper roll-off.

[I wanted a rude username]

I'm not sure 10 uF ceramics were so commonly used at the time the LD1117 was designed. The datasheet specifically calls for electrolytics ... see the app note in section 7.

Try the improved ripple rejection version suggested in that app note, with a third capacitor ... and if that doesn't work, try a 1 Ohm resistor in series with the output capacitor.

[Zero999]

I'm not sure 10 uF ceramics were so commonly used at the time the LD1117 was designed. The datasheet specifically calls for electrolytics ... see the app note in section 7.

Try the improved ripple rejection version suggested in that app note, with a third capacitor ... and if that doesn't work, try a 1 Ohm resistor in series with the output capacitor.

That's a fair point. A ferrite bead in series with the capacitor might also help.

Also try probing the 12V input to see if it's just as noisy or worse.

[wraper]

Yeah MLCC

Then you either need to replace capacitor with different type, add series resistor to increase capacitor ESR or use different IC which is MLCC tolerant.

[wraper]

A ferrite bead in series with the capacitor might also help.

One of worst fixes possible. Not only it negates capacitor effect at high frequencies it also does not increase ESR by known amount at lower frequencies. If you want to add anything, add series resistor.

[wraper]

8.2.2.1.3 Output Capacitor
The output capacitor is critical in maintaining regulator stability, and must meet the required conditions for both
minimum amount of capacitance and equivalent series resistance (ESR). The minimum output capacitance
required by the LM1117 is 10 µF, if a tantalum capacitor is used. Any increase of the output capacitance will
merely improve the loop stability and transient response. The ESR of the output capacitor should range between
0.3 Ω to 22 Ω. In the case of the adjustable regulator, when the CADJ is used, a larger output capacitance (22-µF
tantalum) is required.

http://www.ti.com/lit/ds/symlink/lm1117.pdf

[james_s]

Remember also that a 10uF mlcc is going to give you considerably less than 10uF when you have DC across it. This characteristic is not as widely known as it should be and many of us have been bit by it.