it is better to provide some small electronic circuit with grounding techenics

I have understood a little bit more about this theme. I would like to know if the analog ground is the same than the digital ground. I have a current transformer which I want to measure the analog signal, first it gets into a differential amplifier, then it is rectified and then it is filtered to have a DC signal. This signal goes into the ADC of a microcontroller , PIC16F870. the design: http://my.netomat.net/sardineta/Current_Transformer/ What you recommend me to do? am I right? Happy New Year.

It is very good on EMI and RFI. It's help me to design and in many ways.

It will be nice if all the color illustrations are carried out in the printed edition of the magazine. One can then print out the stuff on the web in a B/W laser printer.

Please make your articles available in PDF as other online magazines do.

I do all my reading while commuting while riding public transport to and from work every day. Having to access the Internet to read your articles forces me to print them out. What I find disconcerting is that you have figures on different pages when you can easily place several on one page, as in the case of the figures in artical 5944. That way less paper trees are used. Of course, someone there may have stock in the paper companies ...

Thanks for listening!

How about split plan with power routing? Will it give better noise and low EMI?

Provide a pdf version of published articles for personal and local archiving. It also makes it easier to reference figures, tables, schematics, etc..

Good article, but please embed the figures. A pdf would be even better.

Provide a PDF link for the DRILL DEEPER articles.

A good article but perhaps a bit empirical on the antenna theory. For example, the author states: "Radiation is, therefore, a function of the loop area outlined by the current path. It doesn’t matter whether one or both of these conductors is a piece of wire, a pc-board trace, or a plane". This is not necessarily the case. A pc-board trace placed above a ground plane creates an antenna having twice the "current-area" (length X current), as compared to the same trace driven against another trace. The radiation from the "ground plane" antenna is 6 dB higher, given the same antenna current. The radiation resistance is 4X for the ground plane case. This is due to the "image" antenna within the ground plane. For an electrically short trace, the vias are the real antenna.

Another generalization implies that the radiation is proportional to the loop area. It is not stated whether the resulting electric field or the resulting radiated energy is proportional to the loop area. Given a square, or a circular small loop antenna (<1/10 wavelength circumference), the radiated power is proportional to the area. However, given a very rectangular loop antenna, such as a microstrip trace above a ground plane, the electric field will be proportional to the via height and not the horizontal trace length. This is due to radiation from the horizontal trace being largely canceled by the magnetic-field cancellation from the ground plane. The vias, being far apart, do not exhibit such high magnetic-field cancellation, and radiate rather well as two very short dipoles.

It is also stated that proper termination can reduce EM radiation. EM radiation is due to the RF current flowing in the "antenna" portions of the circuit. As stated before, this can be mainly the vias. So, if termination reduces the RF current in the antenna portions, then it will reduce EM radiation. If it doesn't, then there is no reduction in EM radiation.

The article also mentions EM radiation from slot. EM radiation from long, narrow slots is largely from the short ends and can be reduced by making the slot very narrow.

Gees... Please just insert all of the illustrations into the text and put all of the text on one big page... This jumping back and forth from page to page is a foolish waste of time.

Very good article. Clearly written with good detail.

As an electronics engineer who has designed over 100 printed-circuit boards, this was valuable information. I welcome these articles, especially the additional information that the website articles provide.

Thank you.

This may be nit-picking, but the story on skin effect isn't quite correct. The formula given applies to a round conductor. Replacing Pi X diameter with 2 X (w + h) isn't the same thing. It isn't that high frequencies like to flow on the surface, they just hate the center of the conductor. In a flat conductor, the currents crowd to the edges and leave the center of the trace with very little current flow. To optimize current flow at high frequencies, round conductors are best and big conductors are better. This presumes litz wire is not suitable.