Screenshot from a PC board I reviewed a few days ago for a client.
The client was told that the GND tracks were necessary to pass EMC tests.
Notice that this is a reference design, so the client didn’t even doubt that it was going to work.
And that is exactly the problem.
You can’t just trust — you need to verify.
The idea here is that by placing a GND track adjacent to the signal trace, it’s going to “sink” or “shield” EMI.
But that’s not what EM fields care about.
For all they care, you could call that trace Mickey Mouse.
The fallacy of the “electrons flowing through copper” model is what keeps you stuck there.
And it’s also what stops you from understanding EMI control in depth.
To really get it, you need to switch perspective and start visualizing traces as merely the boundaries of the fields.
This also means that when you design the PCB, you change your design approach and start seeing things in terms of displacement current, impedance, electric and magnetic fields, and conduction current.
Then, a picture like that will immediately ring the bells and point you toward inductance — and the problems it will create in terms of EMC and EMI control.
To electromagnetic enlightenment,
Dario
By the way, if you want to feel confident about taking your device for an EMC test — knowing you did things the right way, without having to keep your fingers crossed every time, and actually become the point of contact for EMC and EMI in your company, then check out my trainings and courses at fresuelectronics.com/
Especially today, we introduced the Startup and Enterprise Package for small and large teams, where you can get thousands of €€€ in discounts when you purchase the course for you and your team.
Dario Fresu
Screenshot from a PC board I reviewed a few days ago for a client.
The client was told that the GND tracks were necessary to pass EMC tests.
Notice that this is a reference design, so the client didn’t even doubt that it was going to work.
And that is exactly the problem.
You can’t just trust — you need to verify.
The idea here is that by placing a GND track adjacent to the signal trace, it’s going to “sink” or “shield” EMI.
But that’s not what EM fields care about.
For all they care, you could call that trace Mickey Mouse.
The fallacy of the “electrons flowing through copper” model is what keeps you stuck there.
And it’s also what stops you from understanding EMI control in depth.
To really get it, you need to switch perspective and start visualizing traces as merely the boundaries of the fields.
This also means that when you design the PCB, you change your design approach and start seeing things in terms of displacement current, impedance, electric and magnetic fields, and conduction current.
Then, a picture like that will immediately ring the bells and point you toward inductance — and the problems it will create in terms of EMC and EMI control.
To electromagnetic enlightenment,
Dario
By the way, if you want to feel confident about taking your device for an EMC test — knowing you did things the right way, without having to keep your fingers crossed every time, and actually become the point of contact for EMC and EMI in your company, then check out my trainings and courses at fresuelectronics.com/
Especially today, we introduced the Startup and Enterprise Package for small and large teams, where you can get thousands of €€€ in discounts when you purchase the course for you and your team.
1 week ago | [YT] | 11