FEDEVEL
Platform forum

Susano , 05-28-2026, 08:47 PM

Okay I just spent hours debugging noise on an analog MFB stage that was supposed to be attenuated I tried multiple configurations to no avail but as I was calling it a day and turnning off everything, putting my probs back on the wall
I saw the noise still on the scope not sure if i went crazy or what but what in the heck is this noise infesting my probes and analogs ?.

You can see at the scope the 50 hz mains which I know about
But on top there's 30khz signal that bursts for 2-3ms at twice the mains frequency
Its visible on the probes and my analogs even when everything is turned off
I mean this screenshot is taken while my probes are literally in their socket not attached to anything
Ground clips are floating too so I'm not creating a loop antenna actually clipping the gnd clip to the probe gets the signal to disappear


I'm totally confused and I didnt deal with analog electronics that much maybe I'm missing something obvious ? Any ideas would be much appreciated


Thanks

, 05-28-2026, 08:53 PM

🤖 In the meantime, here’s a response from the [FEDEVEL AI HW Assistant]().
It helps to find relevant resources for your question, but it may make mistakes as it’s still in early development.

This is the response from the AI Assistant:
The 30kHz noise you're seeing is likely environmental electromagnetic interference (EMI) picked up by your probe's ground loop acting as an antenna. Here's the breakdown:

1. **Floating Ground Clips Create Antennas**
When probe ground clips are disconnected, the probe tip and ground lead form a loop antenna (several cm² area). This loop becomes highly efficient at picking up:
- Switching power supply noise (common in 20-150kHz range)
- Digital clock harmonics
- Fluorescent/LED lighting ballast noise

2. **Burst Nature at 100Hz**
The 2-3ms bursts at twice mains frequency suggest:
- Rectified but unfiltered 100Hz ripple from nearby power supplies
- Could be coupling from dimmers or poorly filtered AC/DC converters in adjacent equipment

3. **Why Grounding Helps**
When you connect the ground clip to probe tip:
- Creates a short-circuit path for common-mode noise [Video 8](https://www.youtube.com/watch?v=Hj-HVve1QC4&t=792)
- Collapses the antenna loop area from cm² to mm²
- Shunts high-frequency noise to local ground

4.

, 05-28-2026, 08:53 PM

**Troubleshooting Steps**
a) **Probe Setup**
- Always use shortest ground lead possible (use spring clip instead of long wire)
- Keep probe tip and ground <1cm apart when measuring

b) **Environmental Sources**
- Check nearby LED drivers/phone chargers - unplug suspect devices
- Look for dimmer switches or motor controllers in same circuit

c) **Supply Decoupling**
Even with equipment off, residual coupling can occur through:
- Poorly filtered DC rails (add 10nF ceramic + 1µF tantalum at supply pins)
- Ground plane contamination (ensure single-point grounding) [Video 12](https://www.youtube.com/watch?v=2oA4X3OTz0M&t=2856)

5. **Measurement Validation**
- Use coaxial connections instead of passive probes for critical measurements
- Try battery-powered equipment to isolate ground loops
- Add ferrite beads to probe cables near the tip [Video 18](https://www.youtube.com/watch?v=pxgOM3WpoD8&t=3349)

This is a common issue in analog design - the probes themselves become part of the measurement system. Always validate noise floor with inputs shorted before trusting low-level measurements.

You can continue chatting with the AI Assistant here:
https://fedevel.com/hw-assistant

QDrives , 05-31-2026, 01:04 PM

As the AI already mentioned: LED lighting?

Susano , 05-31-2026, 11:36 PM

I don't think so cause I tried proving that
I tried turning off everything in the room and isolate devices one by one
even with everything turned off the noise was still there riding on the mains sine wave only LED in this case I could think of is the oscilloscope screen backlighting itself ??? I also noticed the signal (both mains and 30khz on top) decreases as I approach the screen not sure how does that work

problem is I was prototyping a 40khz ultrasonic receiving stage on a breadboard and it's pretty hard to get a good Q factor my opamps ended up amplifying this 30khz signal either way so probably I'll just give up and render this as something that can't be done on a breadboard after all

QDrives , 06-01-2026, 12:45 AM

"*...that can't be done on a breadboard after all*" -- you are right about that.