Printer causes brownout, UPS wont switch fast enough

[Red Squirrel]

This is a common issue that I've seen in plenty of environments. Currently my parents have it bad.

Laserjet printer plugged directly into an outlet. UPS with the computer plugged into another outlet on the same AC leg or circuit. No option to plug into another circuit as there is none near.

When the printer fuser kicks in, it causes a brownout on that AC leg but the UPS does not switch fast enough, or when it switches it just switches right back, and does this several times. This causes such a distortion in the AC wave that it causes the PC to shut down. Does not always do this, it's random. I think what happens is because it's not a true AC fail situation the UPS does not react properly. Only way they can print is if they flick the power bar off to force the PC to run off the UPS.

From all these bad shutdowns their PC is now acting up. Shutting down on it's own etc... Before I even fix whatever hardware has gone bad (PSU I'm guessing) I need to figure out a solution to this.

Their UPS is an APC 1000, fairly new. At this point I'm willing to even buy a new UPS. Or is there something else I can buy to fix this problem? Some kind of power filter, capacitor bank?

Worse case scenario I'm thinking a battery charger hooked up to an inverter, with a battery. Heck my dad sells RVs, maybe he'll have some of that stuff at the store he can get cheap. I'll have to ask. Only thing is I'd hate to set something like that up and not properly monitor it. Need to find a decent amp charger though, like at least 30 amps or so.

I'm also weary of blaming the wiring, as it's not that old of a house so it has a proper panel, and the basement was redone maybe 15 years ago. House is 27 years old. So it's all romex wire, proper sizes etc..

Though, that circuit IS also GFCI protected at the bathroom. Could this somehow have an effect? I can't see why though.

 

[paul878]

Those cheap ups are not always online ups, so it takes time for it to switch to battery. Maybe you should try a AC Line conditioners and see if that help you. This must be a very old Laser Printer right? New one are less power hungry.

 

[Red Squirrel]

Oh I doubt it's online, it's just a standard home grade APC. If it was online I'm sure this would not be an issue. I think I paid like 100 bucks for it a few years ago. The printer is actually pretty new though. Their old one did not do this. Seems to be hit and miss, depends how they're made I guess. I'm actually surprised this is not more common as I've seen it happen a lot myself. Even at my house the UPS kicks on/off when I print something, but it's an older Ultra and seems to react better than the APC as I've never had the shut down issue.

I wonder if I should just bite the bullet and buy an online UPS, they're actually not as expensive as I thought.

Found one here:
http://www.tigerdirect.ca/applicatio...0368&CatId=234

 

[aigomorla]

320 dollars for a 1000VA :O

RS dont u use RACKS?
At that price point shouldnt u look at a rack online UPS so u can have more felxiablity in where to store it?

And i have an ultra... it sucks.... it has the same problem as your stating...
However my ultra is the 2000VA one... i only use it for accessories cuz of the problem your listing... when the power went off, it wouldnt come online fast enough to keep the system alive.

 

[Gillbot]

My UPS has no problem covering the brownout when my printers kick on. Maybe you need a better UPS?

 

[bryanl]

When the printer fuser kicks in, it causes a brownout on that AC leg but the UPS does not switch fast enough, or when it switches it just switches right back, and does this several times. This causes such a distortion in the AC wave that it causes the PC to shut down. Does not always do this, it's random. I think what happens is because it's not a true AC fail situation the UPS does not react properly. Only way they can print is if they flick the power bar off to force the PC to run off the UPS.

The computer working normally when disconnected from the AC indicates the problem is caused by either:

1) A surge generated by the laser printer. Both the APC and laser printer contain line filters, but they may not be adequate, in which case switching to a power bar containing a line filter may help. The line filter can instead be connected to the laser printer. A real UPS (yours is probably a switch-over supply) is very unlikely to have this problem. APC normally switches to battery at 90-100VAC, which should be good enough for computer equipment.

2) Bad household wiring. Check the voltage under load with a Killawatt connected to a hair dryer or large vacuum cleaner, something that draws at least 10A. A drop of more than 2-5VAC under heavy load indicates a loose connection, and backwire connectors have a reputation for working loose. No-load voltage can look normal even with very bad connections and grossly undersize wiring.

A GFCI should have no effect since it's normally just a pass-thru, but if it trips it keeps the current turned off until it's manually reset.

 

[Red Squirrel]

320 dollars for a 1000VA :O

RS dont u use RACKS?
At that price point shouldnt u look at a rack online UPS so u can have more felxiablity in where to store it?

And i have an ultra... it sucks.... it has the same problem as your stating...
However my ultra is the 2000VA one... i only use it for accessories cuz of the problem your listing... when the power went off, it wouldnt come online fast enough to keep the system alive.

This is my parent's house. i have a pretty beefy UPS setup at home for my server. The ultra is for my workstation and never had issues other than having to change the battery a few years ago as it would not last long enough. Typical wear and tear. My big one is a smarter unit, when it trips, it stays tripped for several seconds even if the power restores, so it's not affected by brownouts as much. My parent's APC seems to trip back to AC too fast so it ends up tripping back and forth a bunch of times in the same second causing the PC to get unclean AC and shut down.

I know the issue is the printer causing a brown out, I just need a solution to stop it from doing that.

If I recall the voltage drops to around 90 volts when the printer kicks in, but I will need to do some further testing. It's also something that happens for no more than a second or two. If their basement was not finished I would just run a dedicated circuit for the printer and make sure it's on the opposite AC leg but not an option unfortunately.

Come to think of it, is there a way I can make a PSU have a ridiculous hold time, like a few seconds? Maybe I can hook it up to a capacitor bank or something? Is there a standard way of doing that?

I'll be buying them a new PSU since the PC now shuts down randomly on it's own, so I'll start replacing parts till the issue goes away, I figure maybe the PSU is the first thing to try. Probably got fried from all these bad shutdowns.

 

[Rubycon]

A ferroresonant based UPS (or power conditioner) will solve this problem.
The UPS definitely has a transfer time, usually a few ms. PC power supplies are SUPPOSED to have a sufficient hold time but it always doesn't work OR the UPS is at fault. It would be easy to see by scoping the UPS' output.

 

[Red Squirrel]

My dad turned off the printer and the PC no longer shuts down at all, so at least that saves me from troubleshooting a PC issue. It's odd though as it was not doing this before, when idle. Though I wonder if maybe they are getting some power flickers that are causing the printer to cycle then it reheats the fuser. The hydro company has been doing lot of work around the grid and maybe it's affecting them.

I'll have to look into these Ferroresonant UPSes, can you recommend some, or what to look for when buying?

If I can find a 13.5v rectifier/psu that has a decent wattage output like 300+ I can maybe just build them a dual conversion unit.

I should try to see if there's a place I can rent a scope, it definitely would be interesting to see what is going on when the printer kicks in. Before, and after the UPS. Always wanted to play with one of those actually.

 

[bryanl]

If I recall the voltage drops to around 90 volts when the printer kicks in, but I will need to do some further testing. It's also something that happens for no more than a second or two.

And you still don't think it's the house wiring? Measure while running a steady high current load rather than an intermittent one like a laser printer.

Most computer power supplies are rated for a hold up time of 16 ms, or not much more than the typical 8 ms switchover time rating of an APC and only 1% of what you desire.

 

[dave_the_nerd]

What the heck kind of printer do you have? I've run everything up to a Color Laserjet 3700 that turned the room into a sauna, and nary a power droop.

If the printer's networked, I'd take the "easy" way out and put it on a different circuit.

 

[Rubycon]

A ferroresonant based UPS/LC is costly, typically these run about $2k for a 1.5kVA model. You will always have 120VAC 60Hz power with less than 1% THD regardless if the input goes between 85V through 140V! Ferro UPS can run with significantly lower input voltage without their inverters running so the batteries really only are used when there is a blackout, for example.

Some laser printers have a very high inrush current when their fuser heaters fire. It can be seen easily on a power line analyzer (Dranetz), some peaks over 40A at 230VAC 50Hz! It's a very brief draw but it can drop the voltage significantly as well. If you use LED or CFL lighting that's regulated, flicker isn't as apparent as if you use tungsten lamps.

If the house has long runs and the printer is at the end the drop is augmented even more. If the wire is aluminum or copper clad aluminum it will also register a higher drop, however most codes when this wiring was implemented specified a gauge higher for that reason. For example a 15A circuit would use 12AWG and a 20A circuit 10AWG, etc.

 

[Red Squirrel]

Ouch, yeah they wont go for that. I'll stick with a dual conversion UPS like the one I found that is about 300 bucks. Though they told me to hold off on it for now.

Next time I'll go there I'll try the blow dryer test to see. So if the voltage stays very low, it could be a lose connection somewhere?

Edit: Was curious so I tried an experiment at my house. Measured kettle + toaster at same time, which came up to 20 amps. So I measured voltage with it off, 120, with it on, 110. Does this sound normal? I will be able to use this to compare my parent's.

My breaker also should have tripped when I did that, but that's another story. I did not leave it long so maybe it never past the trip curve time. My house is much older wiring than my parent's.

 

[aigomorla]

never ask ruby for part advice.. lol..

u want me to get started on how i ended up with a 900 dollar controller when i actually didnt require it?
And here she still tells me i got the wrong one... i should of gotten the 1700 dollar SAS controller... lol..

 

[VirtualLarry]

Could it be that the printer is not the root problem? Instead, perhaps, an incompatibility between an active PFC PSU and a non-sine UPS? In that the printer triggers the problem, but if the power simply went out, the same thing would happen?

 

[Micrornd]

You might want to try setting the UPS up through the APC software.
If you gave a model # I missed it, but most APC units have adjustable sensitivity and selectable voltage take-over.

I'd try that first, as it may be all you need.
My 3kva and 2.2kva units are set for max sensitivity and earliest over and under voltage pickup and never have a problem with brownouts or blips

 

[westom]

Was curious so I tried an experiment at my house. Measured kettle + toaster at same time, which came up to 20 amps. So I measured voltage with it off, 120, with it on, 110. Does this sound normal? I will be able to use this to compare my parent's.

My breaker also should have tripped when I did that, but that's another story.

Breaker should not have tripped. But that involves concepts irrelevant to the problem.

Your 20 amp load causing 120 volts to drop to 110 volts is a useful number. If properly wired, numbers say a 12 AWG wire is about 150 feet from the breaker box. If that wire is not that long, then you know weak connections exists somewhere between that receptacle and the breaker box.

Another useful tool is an incandescent bulb. When the printer powers up, how much does that bulb dim? 90%? Better quality connects mean that bulb should hardly dim.

As bryanl notes, all supplies must work normally and uninterrupted even with no power for 17 ms. Better supplies will work without power even for something like 50 milliseconds. Unstable power from a UPS should be irrelevant to the supply due to how supplies work uninterrupted even during long (tens of milliseconds) periods of no power. However many third party supplies are missing required parts making them more susceptible to short interruptions.

As rubycon notes, a scope is the tool that will answer this question without doubt. A better supply could make a defective UPS look good. An expensive UPS might change power enough that the defective power supply works better. Either identify the problem with a scope. Or just replace parts (ie UPS, PSU) until something works.

Meanwhile numbers created by that 20 amp load imply not fully connected wires in junction boxes for that circuit. An example of facts based in numbers that can result in suggestions not based in speculation when repeated on the parent's circuit.

 

[Red Squirrel]

You might want to try setting the UPS up through the APC software.
If you gave a model # I missed it, but most APC units have adjustable sensitivity and selectable voltage take-over.

I'd try that first, as it may be all you need.
My 3kva and 2.2kva units are set for max sensitivity and earliest over and under voltage pickup and never have a problem with brownouts or blips

Hmm do you know how I would access this config interface? The UPS software only allows basic abilities like shutting down after a certain period.

The voltage drop test I did was at my own house and only for reference. It was a kitchen circuit which I think has lot of other stuff on it, and the plug was probably the last one in the line. All new plugs though, so I can confirm they are well installed and not back stabbed.

I did not go to my parent's yet but I'll have to do further testing. The UPS DOES work when there is a full power outage, it just seems to hesitate during a brownout situation.

 

[Vertaxis]

As a temporary solution, would it be possible to move the printer to a different room on a different circuit? Or, perhaps, run an extension cord from a different circuit to you printer. This should mitigate the problem until the root cause is found and a better solution is implemented. A total cost of 1 extension cord, reusable.

 

[westom]

The voltage drop test I did was at my own house and only for reference. It was a kitchen circuit which I think has lot of other stuff on it, and the plug was probably the last one in the line. All new plugs though, so I can confirm they are well installed and not back stabbed.

Numbers say a significant (but typically not noticed) voltage drop exists. A receptacle maybe 40 feet of 12 AWG wire from the breaker box will create maybe as much as a 3 volt drop and never more than 5. Your drop was 10 volts. If all connections are good, then what caused another 5 volt drop? An example of why useful answers without speculation always use numbers.

Clearly not a safety problem if using copper wire. But human safety would concern me if aluminum wire was used and voltage dropped over double the expected number.

Same test can be repeated on the parent's circuit. Or monitor an incandescent bulb when the printer powers on since quality connects mean a bulb should hardly dim.

 

[Micrornd]

Hmm do you know how I would access this config interface? The UPS software only allows basic abilities like shutting down after a certain period.

It is dependent on model, if your model is recognized by the PowerChute Business Edition software, then those parameters have a limited changeability.
For my Smart-UPSs it allows the following -

High transfer voltage to be set @ 129, 132, 135 or 138v
Low transfer voltage to be set @ 106, 103, 100 or 97v
It allows sensitivity to power quality to be set as high, medium, low
As well as diagnostics, energy usage, and logging.

 

[Rubycon]

Sensitivity on APC also monitors frequency as well. Any setting higher than low won't work when running on an unregulated generator, for example. The UPS will constantly switch back and forth between line and inverter like a slow turn signal! :biggrin:

 

[Scipio Africanus]

Out of curiosity, what is the exact model of the APC UPS causing trouble? There's more than 2 types of UPSs, there is also the middle ground, line interactive UPSs which have an autoformer that can boost and trim voltage. I'm wonder if this one is a cheapy standby type or line interactive.

 

[Micrornd]

Sensitivity on APC also monitors frequency as well. Any setting higher than low won't work when running on an unregulated generator, for example. The UPS will constantly switch back and forth between line and inverter like a slow turn signal! :biggrin:

Interesting, I was not aware of that.
Hasn't presented a problem here, but then most land-based gensets are regulated to + 2 cycles, unlike some older shipboard stuff.
Do you still use that odd 517v 3ph shipboard?

 

[Rubycon]

Our main bus is 11.6kV, we also have 690V and 450V three phase along with 230/115V single phase.