| FORUM

FEDEVEL
Platform forum

USE DISCOUNT CODE
EXPERT30
TO SAVE $30 USD

Question about PCB grounding

NefretinleKandım , 09-23-2026, 07:02 PM
I ve been wondering about this for a long time. In many PCB designs (especially RF and high-speed boards), mounting holes, shielding cans, and metal parts are directly connected to the ground plane.

Why do we design it this way instead of using a separate return path or keeping them isolated from the signal ground? I'd love to hear the reasoning behind this layout practice.
good days.
, 09-23-2026, 07:03 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:
In RF and high-speed designs, connecting mounting holes and shields directly to the ground plane serves critical electromagnetic purposes:

1. **Shielding Effectiveness**
Metal structures act as extensions of your Faraday cage when properly bonded to ground. By connecting shields/mounts to the main ground plane, you:
- Create continuous RF containment (prevents energy leakage through gaps)
- Avoid creating unintended slot antennas between isolated metal parts
- Maintain consistent reference potential across the system ([Source 5](https://www.youtube.com/watch?v=gHF5JyJF-N4&t=6174), [Source 8](https://www.youtube.com/watch?v=vALt6Sd9vlY&t=2936))

2. **Return Current Management**
High-speed signals require uninterrupted return paths:
- Isolated metal creates inductive loops for return currents
- Shared ground plane provides lowest-impedance path (critical for GHz frequencies)
- "Dirty" and "clean" grounds automatically separate through return current localization ([Source 2](https://www.youtube.com/watch?v=Fj9M2CK2cX0&t=3572), [Source 18](https://www.youtube.com/watch?v=2oA4X3OTz0M&t=529))

3. **Practical Implementation**
- **Mounting holes**: Multi-point ground connection prevents resonances (single-point creates λ/4 antennas). Direct grounding avoids impedance mismatches at board-chassis interfaces ([Source 11](https://www.youtube.com/watch?v=yHn-XOcvJOY&t=446), [Source 13](https://www.youtube.com/watch?v=gHF5JyJF-N4&t=2805))
- **Shields**: 360° connection to ground plane (via gaskets or plated edges) maintains continuous shielding.
, 09-23-2026, 07:03 PM
Partial connections create slot antennas ([Source 8](https://www.youtube.com/watch?v=vALt6Sd9vlY&t=2936))

**Key Tradeoff**: While isolation might seem appealing for noise separation, it creates worse problems through:
- Ground potential differences (ΔV creates common-mode noise)
- Increased loop areas (higher radiation/ susceptibility)
- Resonant structures at RF frequencies ([Source 6](https://www.youtube.com/watch?v=vALt6Sd9vlY&t=1284))

Exception: True galvanic isolation (e.g., medical AC/DC) requires transformer-coupled separation, but this is rare in standard RF/high-speed designs.

You can continue chatting with the AI Assistant here:
https://fedevel.com/hw-assistant
Robert Feranec , 09-24-2026, 06:37 AM
Some time ago I was experimenting with mounting holes and guard ring with options to be disconnected from the board ground or connect it and after many experiments the direct connection was the best. since then I always only used direct connection. For example if these two grounds were not connected for ESD there were problems because ESD would need a very big gap, otherwise it jumped between the two grounds anyway and it made things worse. Also, if I dont want to connect mounting holes to the enclosure, I can always use plastic stand off.
NefretinleKandım , 09-24-2026, 08:35 AM
Thank you for this reply 🙏
Use our interactive Discord forum to reply or ask new questions.
Discord invite
Discord forum link (after invitation)

Didn't find what you were looking for?