Author Topic: Automotive PSU Noise  (Read 386 times)

0 Members and 1 Guest are viewing this topic.

Offline NickFearnleyTopic starter

  • Newbie
  • Posts: 4
  • Country: us
Automotive PSU Noise
« on: January 24, 2025, 11:46:57 pm »
Hi all, I've been playing with some automotive electronics and I need a good PSU design for some of my ideas.  I have been using cheap amazon PSU boards but wanted something I can integrate onto my own designs.  I took inspiration from the schematic from an automotive ECU, so the first part of my design (which includes some protection components, a varistor and a fuse and some protection diodes) is a 5V regulator (LM2940).  The second part of the design includes an NCP1117 to provide me 3.3V for an RP2040 IC.

I created a small PCB and had it made up.  I managed to get the SMD components baked on in my little reflow oven (EZMake oven on Adafruit!).

I haven't tested it with any significant circuitry yet but I have loaded it with a 47ohm resistor and measured the output.  It's MUCH noisier than I expected.  I figured the 3.3V side might not be ideal (as I took that design from the PI Pico reference - not automotive...), but the 5V output is also quite noisy.  It has some ripple but minor, the noise is ONLY when the engine is running.  It's an old VW, so really it's ignition and alternator noise...

I used a handheld scope to measure both outputs.  You can see the 5V rail can glitch between 3 and 7 volts, albeit for only around 0.5uS.  The 3.3V rail can glitch between 2.4 and 4.8 volts for around the same period.  I suspect the same transient causing both.  Not sure they're long enough to cause issues but possible, I have SPI (display) and CANbus planned to hang off the rails.

I have added a large electrolytic cap on the 12V rail to see if that had an effect but it doesn't.

Please see my circuit design, the PCB, and two scope images for the 5V and 3.3V rails.

Very much appreciate guidance to improve this (or stop worrying about it.....).  Elsewhere, I've had suggestion to just give up and buy a buck converter but for me this is a learning experience.
 

Offline Smokey

  • Super Contributor
  • ***
  • Posts: 3307
  • Country: us
  • Not An Expert
Re: Automotive PSU Noise
« Reply #1 on: January 25, 2025, 01:32:45 am »
12V automotive systems are nasty environments, especially older cars.  Expect all sorts of bad stuff.

For your LM2940, look at the "ripple rejection" spec and plots. 

You will probably want to add an input filter if you can't deal with the noise getting through the regulators.
 

Offline ahsrabrifat

  • Regular Contributor
  • *
  • Posts: 103
  • Country: pk
Re: Automotive PSU Noise
« Reply #2 on: January 25, 2025, 07:21:36 am »
You can add a TVS diode across the input to clamp transients. You need robust input filtering.
 

Offline NickFearnleyTopic starter

  • Newbie
  • Posts: 4
  • Country: us
Re: Automotive PSU Noise
« Reply #3 on: January 25, 2025, 03:25:10 pm »
Will do and yes better filtering is a good idea. As mentioned the front end (5v) is almost identical to an existing design for an established ECU, so they must have accepted the transients as acceptable for their microcontroller….
 

Offline NickFearnleyTopic starter

  • Newbie
  • Posts: 4
  • Country: us
Re: Automotive PSU Noise
« Reply #4 on: January 25, 2025, 03:26:46 pm »
Thanks for the TVS tip, correct, and if you look at the schematic a TVS is already included as part of the reference I used for the 5V side.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf