EMI and EMC Testing of LED Drivers

Anything that is switched fast or oscillates rapidly, lengths of circuitboard track which act as aerials at frequencies contained in the switching or oscillating.

Without some idea of your circuitry it is hard to be more specific.

There are plenty of good books and articles on this stuff.

Good luck
Del

Speaking of switching quickly or oscillating be sure to look for ringing. Ringing can cause radiation at unexpected frequencies. Even if there is no ringing remember also that a square wave is composed of an infinite number of odd harmonics which can interact with board run lengths and other components to cause radiated signals at frequencies you also might not expect. These are all problems that can be dealt with once identified. You may want to do a pretest yourself with a spectrum analyzer or an o-scope and some tuned circuits.

Thanks Del,

Great to have as resource. Thanks again

Thanks, would be grateful if you can help with the book details.

Regards,

Raju

Good Multilayer PCB design, contain power on internal planes, for a 6 layer PCB these should be layers 2 and 5, any "hot" tracks (fast switching) and I/O traces should be embedded on the internal layers 3 and 4. Outside layers 1 and 6 for connection of SMT components and innocuous traces, a copper fill connected to ground wouldn't be detrimental here either. Enclose the PCB in a screened metal casing if possible any ventilation should be a series of small round holes away from components likely to radiate, any power leads or input/output leads should have good common mode filtering as close as possible to the point of entry to the casing, inside or out. These cables could also be screened if appropriate for the design.

Hope this is of help...

:)

Multi layer board ...for a LED driver?
Hardly economic I'd have thought, I'm assuming this is a very cost sensitive lighting appliction. (But I reserve the right to be entirely wrong)
Del

Good point well made! But how many LED's?

As you quoted, "Without some idea of your circuitry it is hard to be more specific." I thought I'd get more specific without any idea :P

Lemmy

We have range of LED drivers for EMI / EMC test basic model is for 8 LED in terms of wattage it would be 8 watts to 100 watts driver

Yest this is a verhy cost sensitive lighting

The most common drawback of having outside performance testing done on your circuit for any criteria is you not understanding the cause or meaning of a failure report. I understand that this is precisely why you've asked us to interpret something that has yet to happen. Also my initial reply proves that I can be a pain in the a$$ about semantics, but that's enough self deprecation.

As others here have posted, since we do not know your circuit approach or design, comments cannot be anything but wild guesses. I do understand though that you wish to keep some proprietary rights in order to stay employed. So getting down to the basics, there are two methods of coupling from free space to a circuit trace and vice versa. These methods are electrostatic and electromagnetic coupling. Electrostatic field propagation is generated by the varying voltages present and can be easily shielded by the use of some form of a Faraday shield. Electromagnetic field propagation is generated by the varying current produce. To shield this requires enveloping the entire current loop in a box with a higher magnetic permeability, thus providing an easier path for magnetic fields to close their loops than going through your circuit loop. Often a steel box is not practical, but by minimizing the cross sectional area of the current loop path, a less efficient loop antenna to transmit the field into free space will be constructed. These basics can now also be applied to the sensitivity of your circuitry. Keep in mind if your input circuitry is voltage sensitive (MOSFET) or current sensitive (TTL) and you can then decide if shielding or impedance changes will be useful to improve your performance.

Thanks the guesses are true. Still I could get lot of points. Keep sending info.

Thanks again.