Rodel Regucera wrote: > > Ray, > > Can you explain to me your time delay equation with an example. I thought > this as only needed in the high-speed board. I am thinking of how can I > apply it in a high power switching supply with a pwm freq of around 80 khz. > > Thank you in advance. > > Rodel Regucera Rodel- When you place decoupling caps on a board they have a "sphere of influence" or should I say a "disk of influence" that has a radius of about 1/10 the wavelength of the decaps series resonant frequency. This means that current sinks that are within the 1/10 wavelength distance from the decap in question will be effectively bypassed. The 1/10 number isn't a hard number. 1/8 wavelength might be OK too, the point is that the decap needs to be close in relation to the frequency you are trying to bypass. Also the distance is the distance on the board not in free space. Say you are using a large bulk cap to bypass something in your 80 khz switching power supply. We'll assume its series resonant frequency is somewhere near your switching frequency. 1/10 wavelength at 100khz is around 100 meters in air or 50 meters on FR4. So you could place you decap anywhere on the board and it would be just as effective. Now assume you are decoupling a high frequency processor running at 1 GHz. Say you chose a decap that resonates at 1GHz. 1/10 wavelength at 1GHz is about .03 meters in air or .015 meters on FR4. So you would need to place your decap within about 1.5 cm of the current consumer for it to be effective. These examples are described in the frequency domain, but if you change gears and start thinking about time domain you will be able to see that the same arguments hold except that instead of talking about distances you will be talking about time delays. One commercial tool that is useful for power distribution design is Cadence Specctraquest Power Integrity. When you place decaps on the power planes with that tool, a circle of influence is automatically drawn based on the 1/10 (or 1/8) wavelength criteria to assist the user in visualizing where to place the caps. As others have mentioned, the 1/10 wavelenth criteria is just one of many things that need to be considered when placing decaps. Ray Anderson Sun Microsystems Inc. ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List archives are viewable at: //www.freelists.org/archives/si-list or at our remote archives: http://groups.yahoo.com/group/si-list/messages Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu