I doubt this has much anything to do with the normal waveform itself. Based on the price, that's going to be a dumb contact closure device. Even the voltage from those is unstable in a mild brownout, etc. Basically, not suitable for modern electronics. Most anything with an unregulated DC wallwort should be safe, and lightly loaded pcs on old switching power supplies often survive. Basically, stick your Nat router on the contact closure device and get an AVR unit for anything that matters. I strongly recommend cyberpower over apc. If you can use a serial port unit, I may be willing to cut you a deal on a used 900avr with new, extended runtime batteries (I want a second rack-mount unit). <div class="gmail_quote">On Jul 23, 2012 9:59 AM, "Tom Schouten" <<a href="mailto:tom@zwizwa.be">tom@zwizwa.be</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> Hi list, I recently purchased a UPS unit: APC Back-UPS ES BE550G 8-Outlet 550VA 120V The main use is to bridge circuit breaker triggering since we're close to maximum load on some of the lines in the house with the AC units running. I.e. I need about 5 minutes to pick up a flashlight and head to the basement. It seems to work well for my setup when the power cuts out, however when it turns back on there are glitches. My PC is able to bridge the glitch when I put it to sleep mode, but my IP phone seems to reset or hang and the PC resets when it is running. The unit supports 350W and the max load I've measured is 200W. The glitches happen even with a load of about 40W = just the PC. Any ideas what this might be? Is this a known property of these cheap square-wave approximate sinusoid boxes? Cheers Tom _______________________________________________ linux-user mailing list <a href="mailto:linux-user@egr.msu.edu" target="_blank">linux-user@egr.msu.edu</a> <a href="http://mailman.egr.msu.edu/mailman/listinfo/linux-user" target="_blank">http://mailman.egr.msu.edu/mailman/listinfo/linux-user</a> </blockquote></div>