Failure due to harmonics 1

[RajT]

I am on a project where we are commisioning new installed equipment.
The plant UPS system feeds several 110Vac “Beacons” ( stroboscopic lights used by the PA system) . The unit ratg is about 50VA.

We are finding that these beacons are failing after a few days in service.

The supplier of the beacon has carried out bench testing of the units at upto 135Vac and has found no failures.



The strobe units do have intricate electronic circuitry.

The strobe unit design is also some five years old and thus considered proven.



My thoughts are that could it be that the harmonics given out by the ( though less then 5% max) UPS may have some significance to the sensitive beacon circuitry.

As an experiment, I am thinking of fitting some type of RC network filter blocks on the input term of the beacons to see if it is indeed the harmonics that have any connection to the failures.

Any suggestion appreciated.
Guidence to small "filter blocks" supplier web site would also be appreciated.

 

[VEBill]

Did the supplier bench test the failed units, or a sample of new units? If he tested the failed units and found no failures, then you need to start again. If he didn't bench test the failed units, then he should do so to find the root cause.

5% harmonics doesn't sound like much, but stand-by for better informed advice.

I'll put my $1 bet on: 'bad batch'.

 

[RajT]

Thank you VE1BLL.

The supplier did bench test the failed units at a higher voltage after repairing the failed units. I should have stated that the failure was due to a blown fuse in every case.
The supplier also tested new units at a higher voltage and no faults were found.

 

[BrianG]

What sort of waveform does your UPS deliver? How stable is the voltage it delivers? Does the becaon have a fairly small mains transformer in its power supply (i.e. small iron core)?

Small transformer cores can saturate if the input waveform is not a reasonable sine wave - some UPS outputs can look almost like square waves - or, if the waveform is OK but the core is running near maximum flux density, any UPS voltage output surges or spikes could saturate the core momentarily. If the core saturates this would cause a high primary current to flow and hence blow the fuse.

 

[VEBill]

'5% max' harmonics sounds like a pretty good approximation of a sin wave. Time for a 'scope check of the waveform?

 

[MikeHalloran]

Does the PA system "flash" the power to the beacons?



Mike Halloran
Pembroke Pines, FL, USA

 

[RajT]

The UPS o/p is a good sine wave with distortion (THD) of about 3% nominal.

The input transf appears to be adequately rated at 75VA.

The beacon is self flashing ( via internal fixed timing cct)and needs 110Vac power at all tims.

 

[MikeHalloran]

I was thinking maybe the PA system was set up to flash an incandescent light to get people's attention. I assume a self- flashing strobe might not care for all the transients in that service.

They may not care for continuous duty, either. A recent former employer's customers reported lifetimes in days for 115V strobe beacons in 24/7 service.






Mike Halloran
Pembroke Pines, FL, USA

 

[rbulsara]

RajT:

You said the failure was blown fuse. Is that all? So if you replace the fuse, the strobe works again? If so, I would look to upsize the fuse a bit.

Just curious: what is the strobe doing on a PA system?

 

[Skogsgurra]

"The UPS o/p is a good sine wave with distortion (THD) of about 3% nominal"

Is that really so? Is that spec from the UPS vendor? Some UPS have a terrible waveform with lots of HF components and since THD normally isn't measured above 50th (3000 Hz in the US), you will not see them when doing a THD measurement.

I think that it is time to connect a scope to the line and see what it really looks like. HF is known to produce overcurrents.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[davidbeach]

3% nominal THD into what load? Open circuit, or with a nice resistance?

 

[MikeHalloran]

Were the failed units, by chance, "Firebolt" strobes?



Mike Halloran
Pembroke Pines, FL, USA

 

[dpc]

50VA is a pretty small UPS. The strobes may draw a fairly high current while re-charging. If the UPS is heavily loaded, and producing a lot of high frequency noise, the charging caps on the strobes may be drawing enough current to blow your fuses.

 

[VEBill]

I had assumed that '50VA' is the strobe maximum rating, and that the UPS is some huge thing.

 

[rbulsara]

RajT:

Please respond with more clarifications as to qty, size, rating, model no., etc of the UPS and the strobes. Which fuses are blowing?

dpc:

I am not sure Raj meant 50VA UPS, may be he was referring to rating of the strobe unit?

The original question says "Plant UPS", which could not possibly be 50VA.

 

[dpc]

Yes, I agree that I probably mis-interpreted the 50VA reference. More likely the rating of the strobe light.

 

[tequilasunrise]

I've had some experience with Fireball 24Vdc strobes. It was discovered that the manufacturer actually makes two models - a "regular" model and a low-inrush model. We were having problems with inrush and had to switch to the low-inrush model.

On your model, is the fusing manufacturer-installed? Does it satisfy the manufacturer's own requirements for inrush current? Maybe a slow-blow fuse is worth a try.

 

[MikeHalloran]

The same manufacturer makes "Firebolt" strobes, for which we were paying ~$50 direct from the factory, and "Fireball" strobes, which seemed to be available to us only through distributors at ~$250. Photos and drawings indicate small external differences; I have no idea how or if they are different internally. Two emails to the manufacturer about our problem went unacknowledged and unanswered, and left me frustrated and ready to source offshore.




Mike Halloran
Pembroke Pines, FL, USA

 

[RajT]

All,

Thank you for the feed back. Perhaps I should clarify some misunderstanding.
The beacons (with a xenon strobe tube) are "switched" through the PA system when an annuncement over the PA is to be made( they are installed in areas with some noise so the beacons high lights attention) so as Mike Halloran said they are on a short/intermittent duty only.

The input supply to the beacons is 115Vac which is fed from a plant UPS system which is huge at 120kVA.

The rating of the beacons is 50VA.

The sine wave from the UPS is as good as one can get and is within the IEC tolerances of less then 3%THD.

The beacon input circuitry does not have a transformer. The input supply is connected straight acros a dc rectifier block.

The fuse which is blowing is on the "live" line of the ac input to the rectifier.

On the bench test, the input current was measured at 500ma. The fuse ratg is 8ooma.
The beacons are Ex certified and have been in prduction for some 5 years.

On the bench test, the input supply was raised to 135Vac and no failures were noted.

The possibility of a "bad batch" has not been ruled out but is being discounted at the present as the bench testing seems to suggest all is well.

The uprating of the fuse to say 1A (anti surge type) has been discussed but put on hold as the real root of the problem cannot be identified correctly.

 

[tequilasunrise]

You say you bench tested at a higher (135Vac) voltage, but what about a lower voltage? Fuses open due to high current, not high voltage. For a given amount of power, one would expect higher current requirements when the voltage drops.

Is your UPS system regulating output voltage well?

A (relatively) simple test may be to place an interposing relay between the UPS-supplied power and the strobe. Allow the UPS power to energize a relay coil, then use 120V from a separate source (non-UPS) across a normally open coil on the relay to turn the strobe on. This would at least isolate the UPS power as the source of the problem.

Let us know the outcome.