[hpsdr] [Fwd: Re: Tiny Demeter]

[J.D. Bakker]

At 16:43 +0200 28-07-2008, Jeroen Bastemeijer wrote:
>   Currently I'm looking into so called multiphase DC/DC (buck) converters.

Multiphase converters are good when you need a lot of power output 
from a small package with a highly dynamic load. This works very well 
for modern microprocessors, but isn't a great match for an SDR system.

The fact that you're dealing with several separate switchers means 
that you get more switching transients with much RF content, and each 
switcher has a slightly different signature making it harder to 
digitally eliminate the spurs than for a single switcher. Multiple 
sources of EMI also makes shielding and filtering harder.

>Main question at this moment is the maximum noise the HPSDR system 
>can handle on its power supply lines. I hope Phil (or someone else 
>with a Mercury board) will come up with a number one of these days.

It's not just supply line noise, you also have radiated noise to contend with.

Is there anything in particular that you find unsuitable about the 
approach presented in the EDN article that Mike linked to?

JDB.
[one thing that might be worthwhile is to synchronize the switchers 
with the sampling clock, so that spurs end up in fixed locations 
instead of wandering around]
-- 
LART. 250 MIPS under one Watt. Free hardware design files.
http://www.lartmaker.nl/

 1217267971.0