Voltage sags from N-channel mosfet switching 12V, 3A

Other Hardware Motors, Mechanics, Power and CNC
1

Hi! This is my first post and I'm a software guying trying to grok electronics. I didn't take an EE degree but I like to learn more about this area and wouldn't mind diving deep. I still have gaps in my understanding so please bear with me.

Currently I'm trying to build more fault tolerant electrical systems and I've stuck a new oscilloscope that I've bought a PID heating system that I built and used for some time (mega -> n-channel mosfet -> ceramic heater). It works fine, and I sorted out power dissipation on the FET by figuring out what was the appropriate gate voltage, and fixed floating situations by grounding the gate to source with a 10k resistor.

What I've noticed the oscilloscope is when the MOSFETs are not connected, the 12V power supply gives a clean stable 12V. But the moment I connected by MOSFETs (4x fets switching 4 separate loads at 12V, 3A peak), I get an oscillating supply voltage (is this the right term?) and can't to to notice that the wavelengths were about 2ms which corresponds to ~500hz which is about the default pwm frequency of my mega digital pins.

My question is, is this serious? And how can I mitigate this?

The system: I've got the same power supply attached to a stepper driver, and a couple of sensors, (thermistors, linebreak sensors) and a few current issues is that the linebreak sensors get falsely triggered easily when I use interrupts, so I using a digitalRead polling method now. Also the stepper tends to have jittery 'move by itself' occurrences and I had to resort to using the enable pins to shut it down. I've checked and triple-checked that this is not a code-related issue. This is why I'm looking at electrical noise and trying to troubleshoot the problems.

What I've done:

  1. I've added inline filter caps - 2200uf and 104 across power input to my mosfet board. It seems to help with spikes but not really with the oscillations.
  2. I've tried adding a series 220ohm resistor to gate, didn't seem to reduce much.
  3. A ferrite core at gate didn't work either (1 wind, 2 winds).
  4. I've changed to a new 12V supply - nothing either.

I would welcome any suggestions or input, and I would want to confess that my understanding of inductance is spotty.

EDIT:
My images failed to load but there is the supply when disconnected from the mosfets but on:
Imgur

And when connected to the mosfets:
Imgur

2

Please post a complete schematic, with part and pin identification, in line. Hopefully, you are using logic level MOSFETs.

Image posting guide. Many forum members won't bother to open links to other sites.

More hints on getting the most from this forum in the How to use this forum post.

3

You are looking in all the wrong places. Your power supply is either electronically current limited, or is not able to supply sufficient current based on it's design and components.

Paul

4

The most important thing that you haven't stated is the type of the 12V power supply and its capacity to deliver current. What you describe simply sounds like the power supply being unable to cope with 4 x 3A loads. The normal way to solve it is with a better power supply.

Steve

5

Or its simply that your wiring is too thin - wires have resistance and at low voltages and high currents it becomes significant.