I should add that I have since changed over to a linear and it is working much better. The project has dragged on for 3 years now and I want to get it done so I am not going to go back and redesign the switcher again. Next time I have to design one I will bear all this in mind. Sean On Mon, Nov 3, 2008 at 8:55 AM, Sean Breheny wrote: > Hi Olin, > > Well, the first thing I did was to design the switcher layout to try > to minimize radiated emissions. I also had a copper shield can over > the switcher components, which was connected to ground. However, I am > still learning when it comes to good board layout for switchers so I > suspect that I could have done a better job. I agree with you, though, > that shielding is a secondary measure - one should always start by > minimizing the radiation to begin with. > > My receiver is a general coverage shortwave type covering 0.1 to > 32MHz. The switcher was a National "simple switcher" series, I don't > remember the exact part number. I think it ran at 80kHz if I remember > correctly. Basically I could hear a small band of noise every 80kHz as > I tuned, well up into the HF spectrum. I think that the National part > probably placed such a high emphasis on efficiency that it turned on > its output stage very hard and fast, given that it produced audible > harmonics well above 10MHz. > > The effect on the receiver was not overwhelming but definitely annoying. > > Sean > > > On Mon, Nov 3, 2008 at 7:52 AM, Olin Lathrop wrote: >> Sean Breheny wrote: >>> As a hobby project I have slowly been building a high-performance HF >>> receiver over the last few years. I originally used a switching power >>> supply in it. Granted, I'm only slightly beyond newbie when it comes >>> to switching power supplies, but I have to say that I found it >>> impossible to keep the harmonics from the switcher from making it into >>> the receiver. I tried shielding and massive filtering on the input and >>> output leads to no real avail. I did not try magnetic shielding, >>> however the physical separation between the switcher and the rest of >>> the receiver was such that I would find it hard to believe that there >>> could be direct magnetic coupling. >> >> It sounds like you tried everything except the one thing that should be at >> the top of the list, which is proper grounding and attention to where the >> high frequency loop currents go. This is the most important single thing >> for keeping any sort of switching noise from going elsewhere. >> >> A few years ago I redesigned a receiver that had to decode the information >> from two 434MHz receiver and pass the resulting data along over ethernet. >> The original design used a 6 layer board with the digital section in a can. >> The bottom layer formed one face of the can, and a separate piece of custom >> fabricated metal was soldered over the digital and switching power supply >> section. Needless to say, this was a expensive solution. >> >> My redesigned version replaced the analog data slicers with a dsPIC, used >> two switching power supplies instead of one, and only a 4 layer open board >> with no sheild. However, I paid careful attention to where all the high >> frequency loop currents were from the digital parts and the switching power >> supplies, used one layer for a pervasive ground plane, and used sub ground >> planes for noisy sub-sections. In the end we beat the FCC emissions limit >> by almost 15dB, and got a much lower noise floor from the two RF receivers. >> The net result was that it could pick up transmissions from over twice as >> far and cost 30% less to produce. >> >> >> ******************************************************************** >> Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products >> (978) 742-9014. Gold level PIC consultants since 2000. >> -- >> http://www.piclist.com PIC/SX FAQ & list archive >> View/change your membership options at >> http://mailman.mit.edu/mailman/listinfo/piclist >> > -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist