Good, quasi-square wave UPS?

I have a Belkin 350VA UPS with a quasi square wave output. Its output works ok for my computer, but it doesn't work well all with non switching
power supply load. An incandescent lamp flickers badly. A fluorescent lighting with inductive input(magnetic ballast or commercial grade electronic) causes the ballast and UPS to make a noise that sounds like resonating and flickers very badly. The output waveform is very elastic and it jumps around on the scope screen.
The output waveform on my 300W quasi squarewave inverter is much rigid and it has no problem running non-switching power supply loads.
Is there a quasi-square wave UPS that can at least stably power anything my power inverter can?
By the way, there is a dead obvious difference in the inverter circuit between the two. The power inverter uses a few ferrite core high frequency transformers. The Belkin UPS have a low frequency steel core transformer that is used for both charging and inverter transformer.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
in article snipped-for-privacy@news.verizon.net, AC/DCdude17 at snipped-for-privacy@prontoREMOVETHISmail.com wrote on 9/4/03 4:10 PM:

You didn't specify the waveform very well. You could try using a low pass filter that cuts out harmonics. If the waveform is symmetrical + & -, the first significant harmonic will be the third or 180 Hz. I would think that the criterion will be cost.
There are companies in the business of filtering harmonics.
Bill
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Most iron-core transformer inverters don't work well with inductive or high power factor loads. This isn't a limited to small inverters. Some of the big switching power supplies (Cisco routers and switches) specifically mention that they should not be run from certain classes of inverters because of the inverter's inability to handle the switching supply load. The 40KVA UPS at the network lab I worked in a couple of years ago provided a modified sine wave output. The UPS used 400+ volts of NiMH cells as the power source until the backup diesel generators could be started and take the load.
If you buy better equipment (stepped sine wave or sine wave), then the waveform is better and the ability to handle non-resistive loads improves greatly. Of course, you will pay several times more per watt of output than you do for the simple inverters.
I use small UPS's (including a Belkin 350) for graceful desktop PC shutdown and an APC 180 for the local network (DSL modem, router, switch). The APC provides over an hour of run time for the network equipment. Rechargeable power failure lights provide immediate light during short power outages (plus battery powered lanterns) and oil lamps plus a Coleman lantern handle longer outages.
If really pressed for power, I have a small inverter that can be powered from a 12 volt vehicle.

More about me: http://thelabwiz.home.mindspring.com / VB3 source code: http://thelabwiz.home.mindspring.com/vbsource.html VB6 source code: http://thelabwiz.home.mindspring.com/vb6source.html VB6 - MySQL how to: http://thelabwiz.home.mindspring.com/mysql.html My newest language - NSBasic for the Palm PDA: http://thelabwiz.home.mindspring.com/nsbsource.html Drivers for Pablo graphics tablet and JamCam cameras: http://home.earthlink.net/~mwbt / johnecarter atat mindspring dotdot com. Fix the obvious to reply by email.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
An inverter inherently produces a square wave when chopping up a DC voltage. It is the nature of the beast.s There is a way around this which is inherently more complex/ Produce a couple of square waves in the proper phase and magnitude so that they what is called a quasi square wave.
With a three phase inverter this is a good approach.
But if you want a sinasiodal output the sinasoind must be seperated from the square wave by an inductance.
A sinasoid can be produced by using a tuned circuit in the output, which is seperated from the square wave by an inductance.
To produce a sinasiod with 5% , in general the output filter needs a Q of 5.
This means you have 5 times the output current flowing in the tank circuit as flowing out of the output.
With a small third harmonic filter.
In general on all the sine wave converters I have produced I have a tuned output tranformer where the primary is pumped by a pair of transistors and the output transforer is tuned to the output frequency.
Such converters can, if properly designed work into any power factor from 0 to 1 to 0.
The models I have produced are both voltage controlled and current limited.
When the max output current is reached the converters notch back to reduce the output voltage.
Because of the inductor design, the inductor feeds the output transformer 0 volts out is reached well before 0 pulse with width is reached. . . I DO NOT FOLLOW MANY OF THESE NEWS GROUPS To answere me address mail to snipped-for-privacy@aol.com
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Interesting problem. My Belkin 350VA has a slight intermittent flicker in the first 20 seconds or so of powering a 40 watt incandescent bulb (only available lamp for testing - the UPS is w-a-y under the desk), but then the power is stable.
More about me: http://thelabwiz.home.mindspring.com / VB3 source code: http://thelabwiz.home.mindspring.com/vbsource.html VB6 source code: http://thelabwiz.home.mindspring.com/vb6source.html VB6 - MySQL how to: http://thelabwiz.home.mindspring.com/mysql.html My newest language - NSBasic for the Palm PDA: http://thelabwiz.home.mindspring.com/nsbsource.html Drivers for Pablo graphics tablet and JamCam cameras: http://home.earthlink.net/~mwbt / johnecarter atat mindspring dotdot com. Fix the obvious to reply by email.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I do not know about that. At Behlman INvar we produced 500 va power supplies that could go from 0-1-0 power factor and could light a 500 watt light bulb load which starts out as about a 2500 watt load.
It would take a second or so. Initially the supply would go into over load and as the filaments of the bulb lit up and their resistance came up to par the unit would come out of over load and light the bulbs just fine. . . I DO NOT FOLLOW MANY OF THESE NEWS GROUPS To answere me address mail to snipped-for-privacy@aol.com
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.