Extending lamp life of par64 lamps by dimming

[timsabre]

I've kind of got roped into someone else's problem. They have created a Christmas display in a shopping centre and it is lit 24/7 by some permanently installed par cans and, bizarrely, raylights. The par lamps are blowing at an alarming rate. I was surprised to find that par 64's are only rated for 300hrs and raylight lamps only 75hrs.

 

I suggested (after replacing the lamps) that they dim them to extend the life, but does anyone have any experience of how much this will extend the lamp life?

(there is a further bit of fun that several channels of the installed dimmers have got failed triacs and won't dim, but that's a secondary problem...)

 

 

edit to add: I know about the halogen cycle so let's not get into that... but my limited experience of running things 24/7 is that say dimming to 75% would double the life... or something like that... which might just about keep the pars working to christmas.

[Robin D]

Pay a bit more and you can get 2000 hour Par64. Here for example. Even Thomann do them here.

 

Edit: Additional link correctly inserted.

[Brian]

Typically, for a 1k lamp, reducing the applied voltage to 95% of the rated value will double the lamp life and drop the light output to 80%. 85% voltage gives you triple lamp life at 60% light output.

 

GE do make a discharge 1kW PAR64 rated at 3500 hours which are staggeringly bright. A quick look shows it puts out 1,350,000 candelas, or about 10 times as much as a filament version.

 

 

[E2A]

It also appears they do a 150W MH version with a 8,000 hour life.

 

 

[EA2A]

They also have 1k filament versions rated at 4000 hours in 120v (two in series?).

[timsabre]

Typically, for a 1k lamp, reducing the applied voltage to 95% of the rated value will double the lamp life and drop the light output to 80%. 85% voltage gives you triple lamp life at 60% light output.

 

Does dimming count as reducing the voltage? Cos at 80% the peak mains voltage would still be going to the lamp.

There is plenty of light level and they would easily get away with running it at 75%. If the dimmers can get fixed.

 

The pars belong to the shopping centre so I don't think that discharge pars are an option. The long life lamps you're looking at are 500W versions, which naturally have longer life.

[musht]

Raylights can be prone to eating their lamp sockets as well as pathetic lamp life.

[ianknight]

I remember rarely replacing PARs when I was a chiefy but then I used to cheat and set a topset on the racks to 95% and a 5% preheat on the basis that hitting the filament from cold always causes 'shock'. I also set the 'On' level on the desk to 95% on the basis that most of the rig was 1kW Fresnels or PAR's and most of our customers couldn't tell the difference in that last 7-10% - they just wanted it "bright".

 

I remember having a similar discussion with a well known shopping centre in Birmingham along the same lines and Lighthing Technology Projects looked at it - their advice was "Don't use PAR's use discharge and if they really want the PAR can 'look' then sell them a conversion to discharge 'kit'". Eventually the shopping centre stopped talking to me and found an electrical wholesaler who sold them "Long Life" PAR's but from what I gather they only gained about 50 hours per lamp. B-)

[adam2]

Yes, diming most certainly does count as reducing the voltage on a lamp.

 

The RMS voltage is reduced by a dimmer and as a result the filament runs at a lower temperature, the tungsten evaporates more slowly and the lamp lasts longer. Even a slight voltage reduction substantially extends lamp life but unfortunately light output is also substantially reduced.

 

As others post, longer life PAR lamps are available, but remember that they tend to have lower output.

 

Discharge lamps tend to be much longer lasting, and also use about one quarter of the energy of a filament lamp. Electricity is increasingly expensive, and this cost should be considered for long hour display use as distinct from fairly brief stage performances. A 1,000 watt PAR lamp burnt for 1000 hours will cost about £150 in electricity, plus perhaps several lamps. Use of a suitable discharge source would save over £100 per lamp, and the discharge lamp would still be nearly new after 1000 hours.

 

 

 

[dbuckley]

Lamp life varies by the 12^th power of applied voltage, so dimming a bit will significantly increase lamp life.

 

(Yes, not a typo - google it! They worked this stuff out about a century ago)

[Brian]

Lamp life varies by the 12^th power of applied voltage...

 

 

Let's see... 0.95^12 = 0.54. So my 95% and double figure is pretty good.

[adam2]

Yes, this is an inherent drawback of filament lamps. They really need to be operated at within a few percent of rated voltage in order to achieve reasonable life and light output. Not readily achieved in a world in which the mains voltage at the intake can vary by +10% to -6% and by even more than that at the point of use.

 

Use of a 240 volt lamp on a 252 volt supply will roughly halve the lamp life, but 252 volts is within the normal range.

 

Use of 240 volt lamp on a 210 volt supply will substantially reduce the light output, but 210 volts is within the expected range at the point of use.

 

In years gone by, it was common to deliberately run lamps on the "wrong" voltage, either to extend the lamp life if replacement was difficult, and electricity cheap, or in other cases to overrun lamps when the most light from limited energy was more important than the cost of lamps.

 

 

 

[Don Allen]

I still use a few par64, they are two 110V in series as they give four times the life of an equivilant 240V version. I was told it was because there is less stress across the filament and they have more filament mass.

 

doing the maths - 240V 1000W 57.6 ohms vs 110V 1000W 14.4 ohms. larger mass of filament to get a lower resistance.

 

On reflection, I know I got extended lamp hours even though they were always running over volts. One example of the thicker story PARCAN lamp sizes

[adam2]

Other factors being equal, lower voltage lamps tend to be more efficient than higher voltage ones, or alternatively the lamp designer might to choose to design a lower voltage lamp for the same efficiency as a higher voltage one, in which case it will have a longer life.

 

In theory it would make economic sense to run 120 volt lamps in series pairs from 240 volt mains, in practice it is seldom worth the trouble and risks of mistakes.

 

In the case of small lamps, lower voltage are easier and cheaper to manufacture, a 6 volt 3 watt lamp is a lot cheaper and either more efficient or longer lasting than a 240 volt 3 watt lamp. Drifting a LITTLE O/T, In years gone by it was common practice to use 6 volt lamps from a suitable transformer to provide low level night lighting in hotels, nursing homes and similar places. Despite the cost of the transformer and the losses therein, the energy saving and the lower costs of the lamps made this worthwhile.

 

 

 

[KevinE]

I do know that the 'lamp economy' switch on the old NJD datamoons (which just selected a lower secondary tapping on the transformer) extended the life of an A1-259 from a couple of gigs to several months. Phase angle control can sometimes make lamp filaments buzz at very low brightness levels which I suppose is worrying as it suggests the filament is vibrating...though I suspect the advantages of dimming down a bit from full brightness will outweigh any other disadvantage and you can't hear them buzz at brighter levels.

 

I'd have though the major factor on lamp life is the quality of the lamp in the first place and how it's treated. Philips Lamps I think had a table of lamp life vs. voltage reduction on their website. <_<