Look up the generator specs. See if that UPS can compensate for worst-case scenario from the generator. I'd imagine that just about any UPS should do the job, unless it's a real crappy generator. We have a 250 kW diesel generator to power the entire building, and even that causes a noticeable flicker in the fluorescent bulbs. But the computers work fine with the UPS's during the generator startup time and while on generator power. (Various APC models.)

I've got no problems with a cheap $80 APC UPS on my generator. As HRAJOTTE says, computer switching power supplies can take pretty crappy power. They just run it into a diode anyhow and adjust for a wide range of voltages in the output regulation.

Thank you everyone for your input.
From what I have read here and other places, though there isn't a great deal of information to be had on the subject, most opinions opt for the use of a UPS device, to handle stray voltage spikes etc.
I will go along with that.
Thanks again.

A UPS can provide filtering, but you have to watch the specs of the units very carefully.

Many are a straight pass through as long as the voltage in is above ~120 V.

More expensive units have filtering added.

The most expensive units are 'always on.'
They turn the incoming AC into DC for battery charging, and at the same time use generate an output 120 V waveform.
These units usually have the heaviest filtering.
The input AC is never available on the output directly.

One of my job sites resides on a VA campus. The VA is doing "upgrades" and periodically needs to shut off the power to our small building. The other day they shut off the power without notice on a Saturday (we are open from 6-10:00a.m.) and were without power to our building, most importantly, our computer system.

I didn't see this myself, but, when they realized that only powering our lights through their generator wasn't enough, they attempted to plug our APC 1400 UPS' (3 of them, ~1/2 load per)that our server and other network related equipment is plugged into. When they did this, the UPS went berserk, all the lights were flashing, it was making its alarm noise and the office manager told them to quit. Supposedly they checked the generator output and it was 120v. I have no idea what kind of generator it was.

Seems to me, the generator was outputting "dirty" voltage, i.e. wild fluctuations in voltage. This generator exercise took out a linux firewall and a Dell gigabit switch. Once the house power was restored, the UPS' acted normally, re-charged normally and passed all tests.

The VA folks obviously are know-nothings, but, I am very interested in finding an alternative to this situation, both for practical purposes (a power outage will probably happen again) and because it is a fascinating technical issue I would like to solve.

So, what do you guys think??

How old are the batteries? The UPS units themselves?

Have you asked APC about what could make the unit do that? Perhaps the APC support forums might be a better place than here.

Personally, I use a CyberPower UPS unit. I did research and picked that over APC. I work for Hewlett-Packard and have several computers on them in my home office.

@jreagan

The batteries are less than a year old. APC said it is hard to determine, though, given the data, the generator power caused the UPS to go crazy, house power didn't and the unit checks out 100% ok when charged after the power issue. They said it was probably the generator's fault, but, without being able to time travel back and test the output of the generator real-time, it's just guess work.

I have read that if one doesn't go yamaha or honda, most others provide power not suitable for computer use.

"For computer use, require a generator that utilises AVR (automatic voltage regulation) such as the Briggs Pro Max 6000A or a Inverter such as the Honda EU & EM Range"

True? False?

The generator may have been producing a stepped, triangular or square wave rather than a sine wave. Light bulbs don't care but the UPS may not have known what to do with it.

UPSs are somewhat sensitive to the input voltage. The voltage needs to be within a certain range (adjustable on better units) else the UPS will switch over to battery mode. If the voltage is continually fluctuating, the UPS won't like it. But I would like to think that your generator was more stable than that.

"UPSs are somewhat sensitive to the input voltage."

Especially rthe lower price models that just feed the inut through with a little filtering until it falls out of spec.

You are just seeing the input power with limited filtering (at best).

The 'online' units convert the input to DC ALL the time, then charge the battery and invert the DC back to the AC output.

They ARE more expensive, but the battery 'in the line' provides a huge sink for 'noise'.

So I suppose an answer to my question is impossible unless a power outage happens again, the VA tries their generator and I get someone to nte the make/model of the generator so I can determine if it is a piece of garbage or not?

The cheaper the alternator (the power source of every 'generator' you buy) and the cheaper the UPS the less capable they become.

You get what you pay for to a very great extent.

The ratings being slapped on 'generators' are just BS, and the wave shape of most is very poor and gets markedly worse beyond about 50% of any 'rating.'

Purchasing, installing, testing, and checking out a system that will meet the occasional odd-ball voltage delivered by the POCO is not a simple or inexpensive task.

There is a real reason the 'pro designed & installed' systems at major places (like banks) are both large and expensive.

I have no experience with this, but I'd guess it would be cheaper (and safer for the power supply) to charge the PC from your car with a 12v adapter.

Using a generator for a few lights is also overkill - if you're not powering a fridge or HVAC, you'd probably be happier (and would save money) with battery powered lights.

tjdabomb, I have been reading this resurrected post with great interest, Re, generators and UPS power supply for computer use.

The instructions for the UPS I own warn against using generator power as an input into the unit.

That being said,I did a little search using the APC 1400 UPS description from your original post and came up with the user manual linked below.
Look under the paragraph "Voltage Sensitivity" and it may explain why your UPS units were bumping in and out of battery mode while powered by the generator??
I hope this may help.

Here is a link that might be useful: APS 1400 UPS Information.

It would be interesting to check the output of my portable generators on an oscilloscope. While I claim no expertise, it seems that a basic alternator should produce a normal sine wave varying only in amplitude and frequency directly related to the speed of the driving engine. How could the alternator produce square wave AC?
I know that Honda makes some generators that use inverters to regulate the voltage and frequency with varying engine speeds, but most portable generators are not so sophisticated.
I suspect that the UPS may be more sensitive to variations of frequency and voltage in their supply.
I am eager to be enlightened about this.

" How could the alternator produce square wave AC? "

Badly designed.

This often shows up when high 'rating' is more important than power quality (decent sine waves).

The most common cause is inadequate iron in the rotor so that the iron's magnetic field saturates.

You know, looking the the link to the manual supplied by horton, it appears from the operating instructions that the UPS displays would have probably answered all the questions if somebody actually looked at them and took the time to understand what they meant.

yosemitebill -

As my original post stated, I wasn't onsite when someone attempted to plug our UPS' to the generators power - all I know is what a frazzled office manager with no technical experience relayed to me. From the sound of it, it wasn't doing a typical light-flashing show that could be cross-referenced to any manual, this sucker was screaming and the lights were flashing in no discernable order.

Since there seems to be some interest about this subject, I have posted another link at the bottom, to another forum where a discussion took place Re,generators, UPS, Etc.

Here is a link that might be useful: Another discussion on this subject.

While newer PCs contain mandated power factor correction circuitry, its presence tends to make them MORE sensitive to poor waveform quality.

The correction circuitry will try to improve the power factor against the poor waveform, and at the very lest will run hotter.

It may run hot enough to damage itself or fail.

This thread prompted me to search for more information about the power quality from generators. I get the picture and it is not pretty. It could be that two motives, both noble from my perspective, have contributed to this. One would be the quest for low weight for portable generators. My 4400 watt 8 HP (Briggs engine) unit new in 1989 has always performed well for every need. Probably all made in USA. But it is by no means lightweight and I foresee a time when lifting it into a truck by myself will not be possible. The other would be to keep the purchase price low. I can buy today new generators of the same claimed capacity for about the same price as in 1989. So the real cost of producing them has really declined.

I just bought a mechanic special 4400 watt homelite 8hp briggs Bus Driver for 75.00 and fixed the float and it runs like a champ. Its heavy and loud but I have a feeling it will out perform a newer similar rated gen set. I would like to get into the 7500 to 10kw genset range but couldn't pass this up. I might sell this and upgrade but I am not so sure a new 7500 watt unit will perform much better.

It seems the general consensus is that inexpensive generators provide either a modified/square/poor sine wave. My question is, why?? Is it because the alternators are the culprit, and if so, why?? What physical characteristic makes an alternator on a cheap generator produce a poor sine wave?

"My question is, why?? Is it because the alternators are the culprit, and if so, why?? What physical characteristic makes an alternator on a cheap generator produce a poor sine wave?"

Not enough iron in the rotor, inadequate iron in the stator.

If the magnetic field saturate the iron things are no longer linear and you do not get smooth sine waves.

Brickey, so it's not necessarily an issue of the "build" of interacting components that constitute an alternator nor is the issue a matter os quality workmanship necessarily, these units output poor sine waves simply because at least one core element(s), iron, is poor?

Interesting.

Perhaps the rotors and stators are produced in a country where high iron content metals are unavailable/unattainable/too costly?

Back to my original post, I guess I cannot fault the VA "tech" fully, his testing equipment showed 120V but his testing equipment wasn't sophisticated enough to analyze the sine wave that was being output. Though, he wasn't open to the possibility that some other factor, like the sine wave of his voltage, was inadequate.

I love learning this stuff. Thanks all that have contributed!!!

"so it's not necessarily an issue of the "build" of interacting components that constitute an alternator nor is the issue a matter os quality workmanship necessarily, these units output poor sine waves simply because at least one core element(s), iron, is poor?"

The design does not have adequate iron for the magnetic fields being supported.
It is a BAD (and CHEAP) design.

Alternators (and AC electric motors) depend on a careful balance between having enough iron in their pole pieces, the shape and clearance of the pole pieces to each other.

The iron must not become over-saturated by the magnetic field created.

Iron can only contain a certain magnetic field strength before is saturates and the high permeability of the iron reverts to the low value of free space.

There are other metals and combinations that are superior to iron, but they come at very high price premiums and are restricted by this to special applications.

excellent, thanks for the knowledge brickey!!

but..
don't most if not all computers us a switching power supply??
these convert the 120Vac sine wave into about 200Vdc..then convert this back to dc at the needed voltages..

so what would happen if this was feed with 120DC instead of AC??yep it would work just fine..think about it..

so why does off freq and low voltage bother computer eq?..
unless its using a linear supply..think old computers..in short it doesn't..check the spec on a laptop power supply..90-250Vac..no switches..would it work on DC..I think so..

-dkenny

"so why does off freq and low voltage bother computer eq?.. "

Because they have mandated power correction circuits in the power supply.

And if you try to feed them DC you are going to take out the rectifier circuit most likely (and possibly the power factor correction circuits).

The diodes are designed to share the load from the AC waveform, not carry DC continuously.

I guess I'll revive this thread, since I'm again facing this question. Back on Xmas 1998, we had an ice storm where I live which knocked out power for about a week. I had recently installed a Colemate Powermate 55KW generator, which I had set up to backfeed into the house. (Fortunately, we had two functioning wood stoves, since temps were hovering around zero.) I didn't have a transfer switch - I just threw my two mains and closed off the branch circuits I couldn't power. With an electric heat pump, there's nothing to be gained anyway. Anyway, I ran my computer constantly off the generator (with a UPS). Now, being more sophisticated, I'm not going to use the new Westinghouse WH7000 that way. I have a Honda EU2000i which will power the computers...

My Homelite 4000 generator, bought new in 1989, powers my computer with no apparent problems. The engine is one of the old-style Briggs with valves in the block ( flathead). Rugged, reliable, loud.