Why do UPSes not trip fast enough when a laser printer kicks in?

[Bonesdad]

The wiring is fairly new in that basement as it got done maybe 15 years ago, so it's all standard romex, proper awg to code, etc...the contractor who did it builds houses as a living so he knew what he was doing when it was done.

that gave me a laugh...

 

[Anteaus]

If your absolutely positive there is nothing wrong with the UPS, then evidence points at a issue with the circut. There is no way a printer pulls enough amps to brownout a circuit unless something is wrong with the printer or the circuit. I've been in rooms with portable air conditioners and space heaters that were pulling enough amps to dim the lights yet the UPS didn't have problems.

Voltage drops are normal as devices power up, but if your voltage is dropping low enough to trigger your UPS I would start looking for a wiring issue. Assuming everything is fine at the breaker panel, I would take a look at the small junction boxes where the lights and/or sockets split from the main run. It's possible that some of the splices have loosened over time causing poor contact or a short.

Don't assume the wiring is ok just because it's somewhat new or was installed by a pro. Stuff falls through the cracks and sloppy work isn't always obvious. Electrical problems can be so subtle that it's easy to be completely oblivious.

 

[imagoon]

It is going to depend on the UPS. Bigger laser printers (like my HP 4050) say thet need at least 8 amps during power on (120V). However the fuser on a laser printer is nothing more than a 280watt incandescent bulb so the in rush current very high but very short. Cheap UPS units don't always handle this well since the VAC can droop low for 10's of ms until the fuser heats up and the load normalized. Basically the voltage is rapidly dropping, the UPS detects the drop meaning voltage is already starting to drop in the computer's PSU, triggers (say 15ms switch time) which drops the connection all together, detects the VAC returning to normal and then switches back. This can result in a low power back to back switch which the caps in the computer PSU can't handle and shutdown. Better UPS units will switch a stay switched for 10-15 seconds before going back to mains even if power looks good.

Basically this inrush current drops VAC very low but only for a really short period (typically only a portion of a wave). Not enough to trip a typical home resistance breaker. GFCI wouldn't care because power on L1 and returning on neutral is equal.

12 Gauge wire on 20Amp circuits can help quite a bit with this issue since 12 gauge can "respond" to the inrush faster due to lower resistance.

 

[Red Squirrel]

It is going to depend on the UPS. Bigger laser printers (like my HP 4050) say thet need at least 8 amps during power on (120V). However the fuser on a laser printer is nothing more than a 280watt incandescent bulb so the in rush current very high but very short. Cheap UPS units don't always handle this well since the VAC can droop low for 10's of ms until the fuser heats up and the load normalized. Basically the voltage is rapidly dropping, the UPS detects the drop meaning voltage is already starting to drop in the computer's PSU, triggers (say 15ms switch time) which drops the connection all together, detects the VAC returning to normal and then switches back. This can result in a low power back to back switch which the caps in the computer PSU can't handle and shutdown. Better UPS units will switch a stay switched for 10-15 seconds before going back to mains even if power looks good.

Basically this inrush current drops VAC very low but only for a really short period (typically only a portion of a wave). Not enough to trip a typical home resistance breaker. GFCI wouldn't care because power on L1 and returning on neutral is equal.

12 Gauge wire on 20Amp circuits can help quite a bit with this issue since 12 gauge can "respond" to the inrush faster due to lower resistance.

I think that's exactly what happens. When I used to live there it happened to me too (though I had the printer that they have now) and I'd always hear the UPS click twice but it was always too late. In my case the computer would just lock up.

Their basement is finished so even if it is a wiring problem there would be no way to fix it. I suppose I could check every plug connection to ensure everything is tight but that's all I can really do.

I guess I'll start with seeing if they want me to install a new PSU and hope the new one has a longer hold time. Sometimes this is in the specs so I can just look for one where the hold time is high.

 

[JeBarr]

If the home has an attic space above the ceiling, you may very well need to go up in to inspect junction points as was already mentioned, but I don't know if you understood where the location of these splits would be.

If you want to figure this out without spending money then will require MANY troubleshooting steps because there are too many variables. Even if you eliminate the PSU and UPS as source of problem, then you still must inspect the breaker panel and the outlet, which it seems you have already done. After that, it's time for someone to inspect the junctions, assuming there are any. Junctions will be found in attic and hopefully not in-wall which would only happen if your parents hired a shady contractor. (yikes!). If everything about the circuit in question checks out (proper term, guage, etc) after visual inspection and no problems are found then the only remaining causes could be the PSU and/or UPS.

My gut tells me your PSU is not up to the task. My gut feeling has been wrong countless times though, so don't hold me to it.

 

[Red Squirrel]

This is a basement and the ceiling is drywall (even though that was against my suggestion at the time it was built... but their house their rules).

The more I think about it though, I really don't think it's a power issue. It's normal for high draw appliances to cause a brown out when they first turn on. The issue is the UPS not tripping fast enough under a brown out condition as opposed to a full black out, where it does trip properly and fast enough.

If you want to test this take a 1800w blow dryer, set it to cold setting and plug it in, then immediately switch it to the highest heat setting. The lights will dim. This is what is causing the computer to shut down because the UPS is not responding fast enough to the "dimming". I need some kind of capacitor bank or something for the PSU. Just not sure what is the cheapest solution. Perhaps a PSU that accepts a wider range of voltages?

 

[JeBarr]

A quality PSU can make all the difference.

For example, I have what I consider to be a great supply in this rig here. It's a Silverstone st50 something er other...been a while since I checked. Anyways, without it being plugged into any sort of UPS or conditioner, during brown out moments from microwaves, vacuums, etc....and even while having an additional PC connected to the same outlet, under load and printer on the same circuit as well also under load...even with all that and the old wiring in home my PC never shuts down. Not ever.

In a black out my primary PC is the very last device in home to lose power.

 

[imagoon]

This is a basement and the ceiling is drywall (even though that was against my suggestion at the time it was built... but their house their rules).

The more I think about it though, I really don't think it's a power issue. It's normal for high draw appliances to cause a brown out when they first turn on. The issue is the UPS not tripping fast enough under a brown out condition as opposed to a full black out, where it does trip properly and fast enough.

If you want to test this take a 1800w blow dryer, set it to cold setting and plug it in, then immediately switch it to the highest heat setting. The lights will dim. This is what is causing the computer to shut down because the UPS is not responding fast enough to the "dimming". I need some kind of capacitor bank or something for the PSU. Just not sure what is the cheapest solution. Perhaps a PSU that accepts a wider range of voltages?

You really just need a UPS that has a switch delay. It is the back to back switches that kill it. The better units hold the load on battery for a minimum cycle time of a few to tens of seconds before falling back to mains. Better PSU with larger caps helps mask the issue.

 

[Red Squirrel]

Hmm my home inverter charger does this. It gives it a couple seconds before switching back. You'd think this would be standard actually but guess it's not. I might just have to try another UPS, and I can always just use theirs for something else. Really if the one I have was not 500 bucks + batteries I'd just get them one of those.

I might see if I can rig something with arduino that detects a brown out and cuts the power completely and keeps it cut. Will force the UPS into battery mode. I can keep trying different UPSes and different PSUs but this would probably be cheaper.

 

[imagoon]

This is once you get past a certain capacity you get a true UPS versus these partial interrupt supplies. A brown out doesn't really have much of an effect when the inverter is doing 480VDC -> 120/208/220 VAC 100% of the time and the VAC -> VDC stage is just topping off the batteries all day.

 

[bryanl]

This is probably not a matter of the UPS taking too long to switch over to battery but the UPS resetting itself from the laser's power surge, again a matter of the laser's line noise getting into it.

 

[Rubycon]

Either purchase a zero transfer UPS (Ferroresonant are one of the best) or put your laser printer on a dedicated 20A circuit.

The heaters are switched on and off rapidly and the inrush current while very momentary, is very high.

Long runs from the service entrance or panel serving your load can cause excessive drop. Aluminum conductors are suspect as well. Connections should be checked especially if the load test drop exceeds what is allowed by code.