UPS (Uninterruptible Power Supply) Adversely Affects WiFi Performance?

[jzchen]

Well, this is gonna have to be "To be continued" because I moved the BQ16 Pro to another room and connected it to an APC battery backup. (Now I have to prepare for class)...

 

[jzchen]

Ordered one of these to try to sort it out. Some of my old UPSes got LiFePO4 replacement batteries instead of SLA (sealed lead acid). Have no idea how that factors in, maybe over voltage?

GOLDENMATE 1000VA/800W UPS Battery Backup and Surge Protector, Backup Battery Power Supply with LiFePO4 Battery, 10 Years Lifespan, 4 Outlets, AVR, LCD Display, Updated Cooling Fan https://www.amazon.com/dp/B0D7VK8DQM?tag=snbforums-20

 

[degrub]

If they are the same voltage across the cells, then doesn't affect anything we are talking about.

 

[Squall Leonhart]

It would be the square wave approximation that introduces high frequency noise.

the noise is so far below the floor of wifi radio there shouldn't be a problem to consider here.

 

[jzchen]

the noise is so far below the floor of wifi radio there shouldn't be a problem to consider here.

I have a sine wave backup on order. Just plugging into the surge outlets causes normal behavior. Do you have any thoughts on why please?

 

[dlandiss]

the noise is so far below the floor of wifi radio there shouldn't be a problem to consider here.

I wish! I have seen the same symptoms here but it did not occur to me what the problem may have been. Within a few feet of the UPS Wi-Fi bandwidth is wrecked. Actually plugged into it, double-wrecked, like reduced from >500Mbps to <20Mbps.

My main UPS is a good APC model, no problems.

 

[degrub]

It may be some other circuitry that is radiating and not the sine wave approximation. Or it could be normal mode noise that the wall wort ac-DC converter is having issues with. Or just the wall wort itself is failing.

 

[jzchen]

It may be some other circuitry that is radiating and not the sine wave approximation. Or it could be normal mode noise that the wall wort ac-DC converter is having issues with. Or just the wall wort itself is failing.

Thank you for the input.

The "wall wort" was tried on the same UPS, then speed tested in each case: battery backup plug, surge only plug. This meant rotating the "wort" 180 degrees and plugging it back in. (Without carefully moving the UPS to exactly the same spot, the noise appears to be coming out through power, not radiated externally from the UPS. The difference in distance could not be more than a couple or a few inches).

Anybody have advice on a (hopefully not very expensive) UPS model to try?

 

[jzchen]

@dlandiss Might you share the APC model please?

 

[degrub]

is the male plug polarized (larger blade on one of the two) or is it a 3 blade male plug on the wall wort ?

 

[jzchen]

Unless it is 3 pin: GT-BE98 Pro; GT-AXE16000; RT-AX86U, the rest I believe are 2 pin and equal size blades: BQ16 Pro; (RT-AX5400 I reused RT-AC68U adapter). They all have some model of battery backup. The RPs get plugged into the wall but they also are non-polar.

 

[Sander_H]

The UPS is plugged into the wall directly. Prior I was using a cheap surge strip, (but I had forgotten to mention).

Could be the surge strip had some filtering circuit that is not present in the UPS. And thus you are getting more noise from the electrical line. Power from the wall can be "polluted" with radio signals and interference from other devices in your home or even neighborhood.

 

[degrub]

Unless it is 3 pin: GT-BE98 Pro; GT-AXE16000; RT-AX86U, the rest I believe are 2 pin and equal size blades: BQ16 Pro; (RT-AX5400 I reused RT-AC68U adapter). They all have some model of battery backup. The RPs get plugged into the wall but they also are non-polar.

Try inserting the wall wort on the battery side, but keep it oriented the same way you had it on the surge only side. Any difference ?

 

[dlandiss]

@dlandiss Might you share the APC model please?

Sure! <https://www.apc.com/us/en/product/download-pdf/BE425M?filename=APC_BackUPS_BE425M.pdf>

Oh-- did you mean the main unit? APB Back-UPS Pro 1500.

 

[jzchen]

Hmmm. I have the wall adapter connected to a polarized extension cord. When I connect it to the new UPS (supposed to be pure sine wave) I get terrible speed. When I connect it directly to the wall outlet it works great. Sadly, something is wrong that just doesn't quite fit....

 

[dlandiss]

Hmmm. I have the wall adapter connected to a polarized extension cord. When I connect it to the new UPS (supposed to be pure sine wave) I get terrible speed. When I connect it directly to the wall outlet it works great. Sadly, something is wrong that just doesn't quite fit....

That does bear further investigation. My (ASUS) router and access points are plugged into high-end APC UPS boxes and I'm getting 660Mbps downloads, so your issue must be triggered elsewhere.

 

[degrub]

Time to replace the wall wort ?

 

[sfx2000]

Did you unplug the UPS from the wall to force it on battery ?

Is the ups plugged into the wall directly or through a plug bar or other device with filtering circuit ?

I was thinking the same thing - as long as the UPS is on the mains power, it's a pass-through - the inverter only comes into play when running off the battery.

The other thing - the AC adapter is typically going to be a switching supply, so even if one puts 50/60Hz square waves, the output should be fairly clean DC at the rated voltage - I suppose one could always put a scope and see if there are spikes being introduced, but considering how switchers work, I doubt that would be an issue...

 

[jzchen]

I have a Saturday morning class from 8 - 12 but tried to tinker with it a little before and after. Part of it was operator error regarding the UPS. I don't think I properly turned it on and my cell phone may have been connected to/seeing (on Net Analyzer for Android) the DECO BE95 or ACER W6 signals. I have finally resolved how to turn the UPS ON and cancel out the U21 error that showed up at one point.

I did notice something peculiar on my phone while I believe I was connected to the BQ16 Pro AND it was connected to the (new) UPS. The signal strength was very strong say -3x dBm, but when running the Ookla Speedtest app speed test there are 3 little numbers that pop up at the bottom. The first one is always usually small, but the second (download) and third (upload) would rise dramatically, coinciding with (very) poor speed/inversely proportional to speed. The router currently sits about 12 ft above ground level, so it is now very far away from the UPS physically. I strongly believe the issue is some sort of noise coming from the UPS (EDIT- power wire). I did notice that some UPSes do some kind of power conditioning even off battery, and I can share that the new UPS shows line voltage as 116 V if that helps any...

The other new to me CyberPower UPS may be making external noise that is affecting WiFi performance. (Basically I swapped the positions of the BQ16 Pro and RT-AX5400). With the RT-AX5400 12 ft up (far away) connected to an APC that looks just like the BE425M shared by @dlandiss I could speed test 9xx Mbps up/down which is what I expect from the 1 Gig AT&T fiber service. In its new location 2 1/2 to 3 ft from the CyberPower I am getting in the 6xx Mbps range so I suspect it is radiating noise externally, (same "wall wort" used in the 9xx Mbps tests, just a lot closer to the UPS guestimate 10 ft/3 m closer).

 

[jzchen]

Time to replace the wall wort ?

It came brand new with the BQ16 Pro (1-pack). I was looking at a whole slew of them that are available on Amazon. (I did ask a few posts up if you thought they might be worth a shot). I may try to order one, thank you.