On 12-12-97 Paul Williams wrote to Day Brown... 
 
 PW> Hi Day Brown, hope you are having a nice day 
 PW>  
 PW> DB> If I *had* to run a computer on an inverter, I'd use an isolation  
 PW> DB>transformer, and hope the hysteresis would damp the ripple in the  
 PW> DB>sin wave from the inverter, and pick up the slack when the SWPS  
 PW> DB>IGBT shuts on/off.  Of course, isolation xfmrs are big, expensive,  
 PW> DB>and heavy.  
 PW>  
 PW>  Dunno about that. I've seen iso-xfrmrs for under $50US with 2KW
I find a 1500watter at 60$... that weighs almost 20 lb. your 2kwatter 
probly run 25, whazzat another 15-20$ shipping. shouldda defined my 
idea of expensive.  Big, being 6X6X6. izzat big? whatever. 
 
As I told Roy, my problem is with the PWM chips in the new designs 
for PFC SMPS and inverters.  An inverter running a computer, or 
anything else, is a tank circuit.  If you only got one PWM chip in 
the tank, you got no problem.  But, if you got one in the inverter 
to monitor the sinwave output, and another someplace else, things 
can get real interesting real quick. 
 
The PWM chip uses feedback to shape the sin wave, and seems to 
think that it is the only one there doing that. if it mistakes a 
blip from the efforts of another in that tank circuit, which has 
it's own idea of what the sin wave should look like, ... 
___ 
 * OFFLINE 1.58 * BEER: its not just for breakfast anymore! 
--- Maximus 3.01
---------------