Random UPS oscilloscope readouts

R

Red Squirrel

[H]F Junkie
Joined
Nov 29, 2009
Messages
9,211
So I bought an oscilloscope... Thought I'd share various UPS outputs from around the house while playing with my new toy. :D

Ultra 1000AP:


(switching to UPS power)




APC Back UPS ES 750:







Tripp Lite APS 750 inverter-charger: (I was expecting better, this is a slightly higher end unit)




(returning to AC power... good feature of this UPS is that it waits several seconds to ensure AC is stable)


Closer look at the wave



APC Back UPS XS1000:






Note: This is not exactly a scientific comparison, just random tests. To make true comparisons I'd need to pull the power at specific times during the AC cycle to see the different ways they react. They also have different loads. The APS 750 has the highest.
 
Yeah, these are all stepped square wave units. I'd wanna see what a Smart-UPS (True Sine wave) and the higher end Cyberpower (Modified Triangle Wave) scope shots look like.

Units like these will mess with some Active PFC PSUs, so your best bet is to avoid them.
 
Yeah would be neat to test one of those for sure. See how decent the wave is.

One thing I should probably test too with my units is different loads to see how it changes the wave form.

I eventually want to upgrade my Tripp Lite setup to a -48v dual conversion system though, like telcos. I'd use 1800w+ pure sine inverters.
 
After watching two new SLIC huts go up in flames from over loading the batteries, I'd say stay away from the newer gear and go with the older lead acid systems.
 
Yikes, what parts failed, the rectifiers themselves from random failure, or were they overcharging the batteries or something and cause the batteries to go? I'd probably look into some hot swap units like Cordex or something as we use that in lot of our COs so may as well use something I know a big company trusts. No idea of the cost though, will look into that when I do want to upgrade.

May even build my own, been reading up on electronics, new side hobby.
 
Red Squirrel said:
Yikes, what parts failed, the rectifiers themselves from random failure, or were they overcharging the batteries or something and cause the batteries to go? I'd probably look into some hot swap units like Cordex or something as we use that in lot of our COs so may as well use something I know a big company trusts. No idea of the cost though, will look into that when I do want to upgrade.
Click to expand...

None of those have failed. They are just low-quality units. Those who are familiar with electronics should know better than to ever use one of these pieces of crap. Those who are not familiar with electronics, as usual, believe marketing bullcrap and buy this junk rather than doing a little research.

I wouldn't call these "stepped," either. These are all just square waves, and noisy ones at that. "Stepped sine" waves would look like a very low-resolution sine wave and they would have more voltage steps than just positive/negative/off (disregarding the ringing). I'm too lazy to double check, but most likely these are all sold with the marketing claim of being "stepped sine." They are definitely not such. They should be sued for false advertising if they are advertised as such.

I'd say the Ultra and Tripp-lite look the worst, but I wouldn't use any of these. (Actually, is it just me or does the duty cycle on the Ultra look different in the 2 pics? Maybe it's just me. The second one looks better; the first one looks like the duty cycle is way off to me.)


OP, thanks for the info. I knew these were all bad units, but it's always useful to know just how bad they are. I feel sorry that you actually spent money on these, though.
 
Last edited:
Red Squirrel said:
Yikes, what parts failed, the rectifiers themselves from random failure, or were they overcharging the batteries or something and cause the batteries to go? I'd probably look into some hot swap units like Cordex or something as we use that in lot of our COs so may as well use something I know a big company trusts. No idea of the cost though, will look into that when I do want to upgrade.

May even build my own, been reading up on electronics, new side hobby.
Click to expand...

The new (new as in 6 years ago that is) SLIC huts I worked with at Verizon had Lipo batteries. Those things went up like candles.
 
Jorona said:
The new (new as in 6 years ago that is) SLIC huts I worked with at Verizon had Lipo batteries. Those things went up like candles.
Click to expand...

Any idea what caused them to do so and/or what company manufactured those batteries?
 
Jorona said:
The new (new as in 6 years ago that is) SLIC huts I worked with at Verizon had Lipo batteries. Those things went up like candles.
Click to expand...

Oh wow I had no idea they did not all use lead acid. Yeah don't know if I'd trust anything but lead acid in my basement setup at 100+ AH scale. Lead acid systems tend to be very stable.

dandragonrage said:
None of those have failed. They are just low-quality units. Those who are familiar with electronics should know better than to ever use one of these pieces of crap. Those who are not familiar with electronics, as usual, believe marketing bullcrap and buy this junk rather than doing a little research.

I wouldn't call these "stepped," either. These are all just square waves, and noisy ones at that. "Stepped sine" waves would look like a very low-resolution sine wave and they would have more voltage steps than just positive/negative/off (disregarding the ringing). I'm too lazy to double check, but most likely these are all sold with the marketing claim of being "stepped sine." They are definitely not such. They should be sued for false advertising if they are advertised as such.

I'd say the Ultra and Tripp-lite look the worst, but I wouldn't use any of these. (Actually, is it just me or does the duty cycle on the Ultra look different in the 2 pics? Maybe it's just me. The second one looks better; the first one looks like the duty cycle is way off to me.)


OP, thanks for the info. I knew these were all bad units, but it's always useful to know just how bad they are. I feel sorry that you actually spent money on these, though.
Click to expand...

Yeah the tripp lite is the most disapointing as this one is advertised as "stepped approximation to a sine wave". I was expecting better.

The ultra duty cycle does look odd too, but it may be like that when it first starts up. I should test it again and scroll further to see if it changes.

The square waves are fine for most equipment though since it gets rectified into pulsed DC then filtered by caps anyway, though the issue I have with most square waves is how there seems to be a lot of time spent at 0 volts.

I should probe around the switching part of a SMPS while on utility power then while on UPS to see if the differences propagate throughout the PSU. The tripp lite is my "big" UPS and holds equipment up for 4ish hours if power goes out.
 
I'll throw in some measurements on my few UPS units when I get around to buying a differential probe. While 120v "shouldn't" be an issue to measure with a normal probe, I would rather not risk an $800 scope (Rigol DS2072A) for a measurement that doesn't need to be done.
 
dandragonrage said:
I'll throw in some measurements on my few UPS units when I get around to buying a differential probe. While 120v "shouldn't" be an issue to measure with a normal probe, I would rather not risk an $800 scope (Rigol DS2072A) for a measurement that doesn't need to be done.
Click to expand...

Yeah my probes can go 1x or 10x but probably want to get some 100x or higher for line voltage especially if I want to probe inside a PSU as the voltages may end up being greater than 120v in some spots. My tests were done with a AC adapter. I only paid 300ish for my scope but still don't want to blow it up. :D I'd probably want to float it to be safe too. Just don't touch anything... lol
 
Red Squirrel said:
The square waves are fine for most equipment though since it gets rectified into pulsed DC then filtered by caps anyway, though the issue I have with most square waves is how there seems to be a lot of time spent at 0 volts.
Click to expand...

You're right that they are okay for most things, but these have MUCH more high frequency content in them than mains voltage, and many power supplies are not designed to handle it. I ran my refrigerator on a really crappy generator during a power outage a few years back and the low-quality power caused even the motor to sound extremely sick (and it will most likely fail several months if not years sooner now than it would have just from the couple days it was on this bad power). If you ever hear weird noises from a PSU when running on such a UPS/generator it is because there is so much high frequency in a square wave itself, not to mention all the ringing that these ones have. In defense of these UPS, though, that generator (while I didn't measure it) definitely had power even dirtier than this.

I'd consider these units fine for... lamps, maybe? Maybe an alarm clock as long as it's not a cheap crappy one that relies on mains frequency for its own internal clock (fortunately there should be very few that do this these days). Honestly I wouldn't even put my cable modem on it (even if it were owned by the cable company and would be replaced for free if it broke). Though I could understand if someone did use it for things like that ;)

The duty cycle is as it is because they need to do that in order to keep the RMS (similar to average for those unaware) voltage at about 120. When mains voltage is low (say 25V), almost all of your power is coming from your PSU capacitors anyway, but it does have some effect, yes.

Having looked at the pictures again before making this reply, I still can't believe these companies get away with advertising these as stepped sine. These are downright not stepped. At all. Not even slightly. Period. Blatant false advertising. You should sue them.
 
Last edited:
Sure make my wonder just how bad a sine wave my supply puts out...well i guess scopes are cheaper than than used to be, but for now its not in the budget (even though id love to have one)
 
You must log in or register to reply here.
Back
Top