UPS Alarms when 3D Printer fires up

[Quickstep192]

I just got an Uninterruptible Power Supply for my PC and 3D printer.

The UPS alarms for about 15 seconds when the 3D printer prepares to print. Once it starts printing the alarm stops.

I assume the current from starting the heating element exceeds the capacity of the UPS.

Is this OK, or do I need a bigger UPS?

 

[mm08822]

Post the specs of each.

Read what each of the alarm triggers are. Does it require any reset or just clears on its own?

 

[CoogarXR]

If it's an APC it should have an indicator LED of what the alarm is about. Sounds like an overload to me.

 

[mike93lx]

If you haven't, i would confirm that your 3d printer even continues running when the power is cut. It may or may not switch over fast enough for the printer to keep running.

If it triggers and keeps running, I bet it's fine if you can deal with the alarm

 

[Quickstep192]

The UPS is 650VA; the max for the printer is 1000W.

It’s seems that it’s a temporary (15 sec) overload situation; wondering if that’s OK.

 

[wyliesdiesels]

The UPS is 650VA; the max for the printer is 1000W.

It’s seems that it’s a temporary (15 sec) overload situation; wondering if that’s OK.

nope. you should never overload a UPS, especially if you want it to last. thats why its sounding an alarm. to alert you to a problem. some UPSs will even shutdown upon overload to protect circuitry and/or batteries.

get a larger UPS. ideally one that can output 150% or more of the size of the load to give adequate runtime for a proper shutdown.

what is the make and model on the UPS?

 

[mm08822]

The UPS is 650VA; the max for the printer is 1000W.

It’s seems that it’s a temporary (15 sec) overload situation; wondering if that’s OK.

Call APC and ask.
Worst case, plug it in elsewhere for heat up and transfer to ups before running job. When it re-starts after plugging into ups, it will already be hot and assume it will draw lower current for shorter time.

 

[mm08822]

nope. you should never overload a UPS, especially if you want it to last. thats why its sounding an alarm. to alert you to a problem

get a larger UPS. ideally one that can output 150% or more of the size of the load to give adequate runtime for a proper shutdown.

Exactly, it's really a mis-match of hardware.

 

[BillK]

The UPS is 650VA; the max for the printer is 1000W.

It’s seems that it’s a temporary (15 sec) overload situation; wondering if that’s OK.

If I am doing the math right 650 VA at 120 volts is 650 watts. If the printer is 1000W then obviously the UPS needs to be larger ? Not sure I would want a 3D printer on your typical UPS. How long will your UPS run at 650 watts output ?

 

[Quickstep192]

I put an ammeter on the printer and monitored during the printing process.

The printer draws 1000 watts only for the time while it’s heating the print head and printing plate. After that, it drops to under 600 watts, but that’s still on the edge of the UPS’ capacity. And that’s with basic PLA filament; some filaments require even hotter temperatures.

I really only wanted to protect the printer against momentary power blips and maybe power it until I get the generator going.

-But-

Either way, to paraphrase the line from vs Jaws, “We’re gonna need a bigger UPS…)

 

[mike93lx]

I really only wanted to protect the printer against momentary power blips and maybe power it until I get the generator going

Again, have you tested this? The ups may not switch over quick enough to prevent the printer from rebooting

 

[wyliesdiesels]

I put an ammeter on the printer and monitored during the printing process.

The printer draws 1000 watts only for the time while it’s heating the print head and printing plate. After that, it drops to under 600 watts, but that’s still on the edge of the UPS’ capacity. And that’s with basic PLA filament; some filaments require even hotter temperatures.

you forgot the load of the PC

I really only wanted to protect the printer against momentary power blips and maybe power it until I get the generator going.

-But-

Either way, to paraphrase the line from vs Jaws, “We’re gonna need a bigger UPS…)

@ more than 600w (including PC) youll get all of a few mins max

 

[ripperd]

Why is your 3d printer on the UPS in the first place?

 

[Quickstep192]

Why is your 3d printer on the UPS in the first place?

I wanted to make sure printing would continue if there was a momentary power blip. We have a lot of blips here.

It would be a shame to be 7 hours into an 8 hour print and have to start over because the print got interrupted.

 

[ripperd]

I wanted to make sure printing would continue if there was a momentary power blip. We have a lot of blips here.

It would be a shame to be 7 hours into an 8 hour print and have to start over because the print got interrupted.

Yup, OK that makes sense. Just making sure you didn't intend for it to be able to finish a job in the event of more than a blip.

As others have said, need to size the UPS to be able to adequately power the system (with a little headroom). 650va definitely sounds a bit light. Given the equipment I think you'd want something 2x that or more.

 

[ripperd]

Also, do you need the PC on the UPS at all? The 3d printer I have doesn't start printing until the PC has sent the entire job. Once printing starts I can shut off the PC that sent the job, it isn't needed anymore.

 

[bwringer]

I wanted to make sure printing would continue if there was a momentary power blip. We have a lot of blips here.

It would be a shame to be 7 hours into an 8 hour print and have to start over because the print got interrupted.

Yup, you're gonna need a bigger UPS. I would absolutely put a 3D printer on a large UPS by itself; one blip in an hours-long print would trash the whole thing.

It does need to be sized to support the largest possible load, even though the printer doesn't use that much after warmup. I'd also think about how long you want it to be able to continue on the UPS; get through a blip, get through 30 minutes, an hour, etc.

And yep, put the PC on a different UPS. I'd also put your internet/networking stuff on a UPS, especially if the printer has a wifi connection; some devices will stop everything and try to reconnect if wifi drops.

In my case, I sized everything so that I should be able to run my monitor, dock, and LED lighting as well for at least an hour, enough to finish a Zoom meeting.

 

[ripperd]

In my case, I sized everything so that I should be able to run my monitor, dock, and LED lighting as well for at least an hour, enough to finish a Zoom meeting.

Unrelated to the OP, but I have a UPS on our fiber adapter, router, and WiFi system, with an upgraded battery. I'm at close to 2 hours runtime. So as long as my laptop was fully charged, I have 2 hours that I can keep using the internet and working nearly as normal, just no external monitors/docking station. I am partial WFH, my wife is 100% WFH, so having good infrastructure there is very helpful.

If it goes longer than that, I can hook up the generator haha.

 

[BurtEggley]

yes, get a new UPS that is preferably 2000 watts. The batteries on many UPS brands (especially APC) are sealed lead acid, and they only give full output for 3 +/- years before the batteries begin to degrade. Some brands are lithium ion. If you get one of the LION ones be sure it is UL approved, and not some off brand with no history, due to continuing fire risk with LION products in general. 1500 watts would work too but 2000 will give you more run time.

 

[wyliesdiesels]

using the apc calculator, inputting 1100w for the load w/ 10% extra capacity and 30mins run time, youre looking at a 1500va UPS with 1 expansion battery. this will cost you close to $2K

if you need longer run time than that, say an hour, youll need 2 expansion batteries.

 

[Innovate1]

If I am doing the math right 650 VA at 120 volts is 650 watts. If the printer is 1000W then obviously the UPS needs to be larger ? Not sure I would want a 3D printer on your typical UPS. How long will your UPS run at 650 watts output ?

Not sure exactly what math you are talking about. There wasn't enough information to do any math. VA = Watts only if the power factor is 1. First thought was that heaters are typically a power factor of 1 since they are just resistive loads. But for a 3d printer they are likely powered from a power supply for safety isolation. So we need to know the power factor of the power supply in the 3d printer. I looked up details for an Ender 3 V2 which is a popular hobby level unit. The power supply has power factor correction for a power factor of 0.95. Thus watts = 0.95 x VA. Some loads have much worse power factor so there is more difference between watts and VA. Max power for the Ender 3 V2 is 335W and about 125W typical when printing. 1000W seems like a lot but it may be a much faster printer.

Really need to size a UPS on VA and not on watts for loading (watts is good for run time). A lot of things now have power factor correction so it's less of an issue than it was but still important.

 

[rlitman]

If I am doing the math right 650 VA at 120 volts is 650 watts. If the printer is 1000W then obviously the UPS needs to be larger ? Not sure I would want a 3D printer on your typical UPS. How long will your UPS run at 650 watts output ?

650 VA kind of equals 650 watts. However, while 3D printer heating elements are resistive (so PF theoretically equals 1 at full power), they're running on DC switching power supplies, so expect a capacitive PF of perhaps as bad as 0.8.
At PF=0.8, 650 VA is only 520 watts. Coincidentally, most UPS are marketed with the VA rating, because the watts are that much lower. I'd bet that when you look at that UPS's nameplate, it also happens to say the output is only 520 watts. Knowing that...

I put an ammeter on the printer and monitored during the printing process.

The printer draws 1000 watts only for the time while it’s heating the print head and printing plate. After that, it drops to under 600 watts, but that’s still on the edge of the UPS’ capacity. And that’s with basic PLA filament; some filaments require even hotter temperatures...

Forget what it drops to. First, because your UPS circuitry needs to be rated to handle the peak load, but also because the printer is NOT dropping to 600 watts. It's using PWM to pulse the 1000 watts in small durations to give a thermal average of 600 watts. The UPS is ALWAYS seeing the 1000 watts.

And knowing that the power factor is not 1, your 1000 watt printer needs a 1250 VA (or larger) UPS.

 

[Innovate1]

650 VA kind of equals 650 watts. However, while 3D printer heating elements are resistive (so PF theoretically equals 1 at full power), they're running on DC switching power supplies, so expect a capacitive PF of perhaps as bad as 0.8.
At PF=0.8, 650 VA is only 520 watts. Coincidentally, most UPS are marketed with the VA rating, because the watts are that much lower. I'd bet that when you look at that UPS's nameplate, it also happens to say the output is only 520 watts. Knowing that...


Forget what it drops to. First, because your UPS circuitry needs to be rated to handle the peak load, but also because the printer is NOT dropping to 600 watts. It's using PWM to pulse the 1000 watts in small durations to give a thermal average of 600 watts. The UPS is ALWAYS seeing the 1000 watts.

And knowing that the power factor is not 1, your 1000 watt printer needs a 1250 VA (or larger) UPS.

I agree and just want to add that PF can be less than 0.8 for simple rectifier and capacitor input power supplies so it could be even a bit worse than your example.

UPS's may use the VA because the number is bigger than watts for marketing but there is a good technical reason to do that too. The switching devices (and other parts) in the inverter output are stressed by current not wattage - 500VA stresses and heats the output stage the same or nearly same (perhaps some minor other effects) regardless of if the actually watts are 500 or the power factor is low and the watts are 300. So a VA rating makes sense even if many in the general public don't know the details of what that means. Run time estimates should be based on watts.

 

[rlitman]

...Run time estimates should be based on watts.

The rest of your post I agree with (and yes, rectifier/filter supplies can have very low PF, though a 3D printer won't have such an old-school design), but I don't think so about watts and run-time estimates. The UPS has to generate apparent power, and the imaginary power component is still lost as heat (within the UPS). Sure, at the battery side of the bus the draw is in watts (it's DC), but the UPS conversion efficiency will vary with the PF, so I'd guess that the battery real power draw more closely matches the UPS apparent power output (AC VA) than the UPS real power output (AC watts).

 

[Innovate1]

The rest of your post I agree with (and yes, rectifier/filter supplies can have very low PF, though a 3D printer won't have such an old-school design), but I don't think so about watts and run-time estimates. The UPS has to generate apparent power, and the imaginary power component is still lost as heat (within the UPS). Sure, at the battery side of the bus the draw is in watts (it's DC), but the UPS conversion efficiency will vary with the PF, so I'd guess that the battery real power draw more closely matches the UPS apparent power output (AC VA) than the UPS real power output (AC watts).

Yes, most electronics these days have power factor correction and a good PF, probably 0.9 or better. So the whole discussion of VA or watts, in most applications, doesn't really give much different results.

But technically UPS run time would be most accurately based on watts. Here's a ups manufacturer run time chart.
https://www.cyberpowersystems.com/tools/runtimes/
Nowhere is VA found. The imaginary power isn't all lost to heat - it goes back and forth between the load and battery/bus caps. The ups will have more losses with a poor pf load for the same watts. But those additional losses are going to be small compared to the load wattage. Consider a power factor of 0.8 with a load of 1000W. VA would be 1000/.8 = 1250W. The load current on the output will be 25% higher than for a PF of 1.
PF of 1 line current for 120V power is 1000/120 = 8.33A RMS
PF of 0.8 line current is 1250/120 = 10.4A RMS
If we estimate the losses in the inverter to be 10% (which is likely more than it is - I design small inverters this size for mass market, cost sensitive applications and they are typically 95 - 98% efficient). Thus the losses for 1000W would be about 100W (worst case and likely much less). If this loss goes up proportional to the current (a reasonable approximation) this would be an extra 25W loss (125W total loss).

Total of inverter and load power (loss) for PF=1 is 1100 W.
Total of inverter and load power (loss) for PF=0.8 is 1125W.
The change is 1125/1100 = 1.023 or a 2.3% increases in losses.
This is probably within the accuracy of the run time estimate and nowhere near the 25% increase that just using VA would give.

Note that there other losses in the UPS front end and battery but those will be affected by PF even less if at all.