CM Chokes sometimes suck

Yeah, if you're wrong when you say to take them out. It's never going to be a 'no choice now' situation, though: stand and fight. The defensive "do EVERYTHING we can" style is bad engineering, bad tactics. Your job is to use planning time and resources to good effect, and project hardware bloat is NOT a good effect.

"No evil, no matter how small, ought to be tolerated." -- Aesop

============ we are talking about a few pennies worth of parts, i'm not sure I would use the word evil.

if the EMI guy is responsible for the EMI compatibility issues and he wants the CM filters in, I would let it go. It's his contribution to the effort, do not minimize it.

You never know, there may be 5 GHz RF fields and the CM filters help.

You never know, some issue, EMI or otherwise, may come up in the future and having those CM filter pads on the PWB may be helpful. It's cheap insurance.

Most importantly, maintaining a good working relationship with other members of the team is worth ,much more then the few cents for the parts. If the customer insists on the EMI guy being part of the team, I would even suggest soliciting the EMI guys input at the start of the next project.

The real issues here are much larger then engineering.

Mark

Clearly Aesop wasn't an engineer.

hired an EMI guru. He insists on placing CM chokes on every single ended I/O signal line. These signals can be floating 100K thermistors that are received by a Pi filter (100K resistor and two 1.0uF caps to ground). The miniture CM chokes are 1k at 100MHz (1.6uH).

Have you killfiled google?

If the level of evil is guaranteed to be substantially less than the specified limit, it's probably overdesigned.

But they do hurt. They cost money. They do not provide the intended value. And worst of all they teach erroneous design techniques to the easily swayed. Costing lots of money permanently.

A far better design would use a tapped inductor 2 turns to the outside world per turn to inside on the two sides of the tap; then place a cap to ground at the tap. Toroids or closed pot cores preferred.

Measure the performance difference in front of management.

?-)

signal is "single ended".

That is a very different case. The CM choke might actually help. Drawing the circuit diagram.

___________@@@@___________> to measuring circuit Therm O=======CMchoke===

Now what's a deecee-bel? Our MBA professors never mentioned ... oh wait, it's a cheese, right? Yes, it is a cheese!

Custom inductors to save money?

Why not a 5-cent dual ferrite bead?