Modulated Sine wave sensitivity?

[sbyrd]

I have an issue where I am adding UPS power to my tower sites. I am using the Tripplite APS750 inverter/charger at most sites with good success.

However at one site I have deployed the slightly smaller Tripplite APS700HF inverter/charger and have an issue where it goes immediatley into an overloaded state as soon as I disconnect from power and switch to battery.

After discussions with Tripplite support they are leaning towards my equipment on the towers having issues with PWM modulated sine wave form power and even though when on the Mains the power draw is only 200W that the equipment is sensitive to PWM and is requesting more power than the inverter can supply (700W) which causes the overload.

Support is basically saying that when on PWM power the equipment is instead of sensing 120V of output power is sensing something like 95V of power and the amperage draw of the equipment is increasing to compensate so it can output the needed watts.

So if the inverter can do 120V at 5.8A (700W) and when on 120V pure AC power my equipment needs 2.1A (250W) then if the equipment is sensitive to PWM and sees the inverter as an undervolt then there is the chance my equipment is now trying to draw more than 5.8A which causes the inverter to sense and overload.

I was told by their support to ask the equipment maker if they know if their equipment is sensitive to PWM modulated sine wave form power or not.

So I ask are the Netonix WS-AC switches sensitive to PWM or modulated sine wave?

[Dave]

well, never heard of this issue before, and I have no idea..I will forward the question to our AC power supply manufacturer....

[Dave]

our AC power supply manufacturer said that their AC supplies will not work correctly with a modulated sine wave source, they are not designed for it.

[sbyrd]

Thanks for the response. Did they give any details on what issues it would cause or just that it wont work?

Reason I ask is I have 4 other sites running the Tripp Lite APS750 which is also modified sine wave and it ran the equipment fine on battery power. However it was only on battery for 5-10 minutes as part of my test. I have not tried to run it longer on battery.

[Dave]

all they said is that the PFC circuit will not work normal & it will use abnormal current....it might be that it will work okay for most of your trip-lite converters, but if one of the trip lights is a bit out of frequency, it might give our AC power supply a headache...sorry, thats the best I can do....

[sbyrd]

I appreciate you looking into that for me.

[sbyrd]

One last question and if you don't have any idea I completely understand.

I guess what I am trying to understand is Tripp lite correct that running Ubnt and Netonix equipment on PWM sine wave power is causing either a true or false overload situation? I am not an electric engineer but I could not find anything online that said that running equipment on PWM power would cause it to draw more load then when running on pure sine wave.

I can understand if the radio/switch equipment runs oddly when on PWM and the power is less stable, but I have a hard time understanding how 200-250W of load on a 700W inverter would go to overload.

If I am explaining what I was told correctly Tripp lite support said that when on PWM power some equipment could sense that as too little voltage so it requests more power to compensate and the inverter sees the request for power would be over the capacity of the inverter so it goes to overload.

He said that on the APS700 that it can only run for 1s in an 125% overload state, but the APS750 can run up to 1hr in a 150% overload state. He said that is why my other sites with the same telecom equipment are working on battery because the APS750 can run overloaded for longer and he thinks that those sites will not run as long on battery as I expected due to that.

[Dave]

for sure, if you feed our AC supplies a modified sine wave source, they think the voltage source is wrong & can use excessive current...lots of different equipment have problems with modified sine wave sources.

the answer really is that our AC supplies (and others) will actually cause error conditions inside the power supply circuits & really use to much current.

[sbyrd]

Unfortunately that is not what I wanted to hear.

I am not an expert on these things and usually used small APC UPS at tower. I replaced them due to the very short run times. Come to find out most of them are pure sine wave. I chose this inverter and setup based off the recommendations in an article on SwiftFox's website about how they add long running backup power to WISP sites.

They did not mention and I never knew or thought to investigate what type of power they output and assumed things would just work.

Now I am concerned for my other sites I moved to the APS750 that they might not run as long as expected or have major issues now due to the PWM power. Plus I have now spent decent money and time on a seemingly bad solution.

[KBrownConsulting]

@Dave,

I stumbled across this topic & it's causing me to be concerned about a proposal I'm putting together for a client (as well as some past projects!) that includes Netonix switches & battery backups to power them. I'm hoping for a little more info on this topic if possible.


First, a couple reference articles I found while trying to research this to be better informed, that have seemed relevant:

• Pure Sine Wave vs. Simulated Sine Wave
• Computers containing PFC (Power Factor Corrected) power supplies and their use Back-UPS and Smart-UPS SC with Step Approximated Sine waves

My takeaways so far:

1. Per Dave's comment in this thread, all Netonix Switches have PFC power supplies in them which "will not work correctly with a modulated sine wave source, they are not designed for it."


2. According to the first article I linked above, I believe that all battery backups that provide a "simulated or modified sine wave, also known as a pulse-width modulated (PWM) sine wave" would fall into the category of "modulated sine wave source" and therefore "will not work correctly" with Netonix switches.


3. Unfortunately it seems that pure sine way UPS devices are much more expensive. I don't have any definitive comparison but a brief search seems to suggest they tend to run at least 2x their simulated/modulated/PWM counterparts.


4. According to the second article, it sounds like PFC power supplies can very briefly draw their max power when initially powered on (period of inrush current) and due to inefficiencies of PFC power supplies this can often be 125% of their rated capacity. When a UPS transitions from AC to battery power, even though it happens incredibly quickly, there's enough of a delay that the power supply thinks it being initially powered on & the inrush current period may be triggered causing the power supply to draw maximum power momentarily.

Questions:

1. Related to statement #2 above, would you agree that all battery backups that provide a "simulated or modified sine wave, also known as a pulse-width modulated (PWM) sine wave" fall into the category of power sources that Netonix switches are "not compatible with"?


2. Related to takeaway #4 above, is the problem just related to the inrush current and if so, would selecting a battery backup that has the capacity to handle the inrush current be enough to mitigate the problem? If so, can you tell us what the maximum inrush current draw is for the power supplies in Netonix switches? Or at least tell us what the power supplies are rated for so we can estimate the inrush current?


3. This last question might be a little harder to answer, but even if the problem is not related to the inrush current and it's actually a problem of the switch drawing a lot of extra current the entire time it's on battery power; according to some of the comments in this thread that dsbyrd relayed from TripLite as well as his own experiences, it would seem that as long as the battery has enough capacity to handle the extra current draw it should still "work", although with (potentially significantly) reduced runtime. Would you agree with that statement? And if so is there any way to know how much extra current extra draw the switches are going to require so we can evaluate if the UPS is sized appropriately? For example I'm curious if the extra draw is a percentage of the switch's total power supply capacity, or a percentage of the active power load on the switch?


4. I guess at the end of the day I'm trying to establish if there's a formula that we can use to reliably estimate whether a battery backup that produces a PWM sine wave will work & what the runtime will be, or if it's a case of needing to always ONLY use pure sine wave battery backups.

Since the answers to these questions greatly impact the cost & reliability of Netonix deployments I'd appreciate any feedback people have!

Thanks