I can't believe I didn’t think of this sooner. For months there has been a debate about of which type of lamp will be brighter short arc or long arc. I have always said that the brightness should not be much greater but some have said that a vary short arc lamp will be twice as bright as bright so who is right?

For this test the front fresnel is the 650mm from LL the triplet is my Buhl 18" fl 128mm diameter lens both lamp images fit 100% into the triplet. My Fresnels alignment is not perfect and you can tell by the lower readings on the bottom of the screen.

For this test I tried to be as accurate as possible I did not move the lamp when adding the reflector I put blue tape on the screen so the reading with the lux meter would be from the same place and most important is the lamp was exactly 333mm from the rear fresnel and exactly centered in the triplet.

I took two sets of LUX readings for each setup. The stray light did vary a lot. The Ikea reflector stopped a lot of stray light from lighting the back half of the room.

The parts

Reflector Ikea
Lamp
1) Long arc was the FDF T6 Venture 13,250 lumens lamp with a 60mm filament
2) Short arc was the FTK gy9.5 OSRAM 12,000 lumens with a 13mm filament

Both lamps are 500 watt but the short filament FTK makes 1250 less lumens

[EDIT] Tested output power has lots of stray light and was hand held.

A reading with the LUX meter at 3' gave 1622 LUX for the short FTK lamp and 1821 for the long FDF lamp

The test screen

Click to view attachment

The first lamp to test is the FDF 60mm filament with no Ikea reflector at 333mm from the Fresnel the screen is 49" X 35"

Triplet focus point shot.

Click to view attachment

The LUX readings

400 614 328
415 758 345
333 586 306
Average =453 LUX

Stray light
60 48 60
49 64 47
48 50 50

Average LUX =52 adjusted LUX =401

A shot of the lamp image with the Ikea reflector offset to let you see both images. The reflector can't focus on the whole image.

Click to view attachment

Now the FDF with the Ikea

590 727 378
519 870 447
449 706 360
Average =561 LUX

Room light
51 29 48
40 47 35
42 27 36
Average = 39 LUX Adjusted LUX =522


Now the short arc lamp the FTK 500 watt 13 mm filament lamps. A picture


Click to view attachment

And the lamp image shot at the triplet

Click to view attachment

The first LUX readings with no reflector at exactly 333mm from the rear Fresnel

404 501 400
428 565 414
355 469 335
Average = 430 LUX

Room light
55 42 53
51 67 42
55 41 43

Average = 49.9 LUX Adjusted LUX = 380.1

Now with the Ikea reflector

539 707 565
600 911 622
538 656 551

Average = 632 LUX

Room light at the screen
44 31 48
34 52 34
34 24 42

Average = 38 LUX Adjusted LUX =594

The next test was when I forgot to put the heat glass back in when changing lamps.

This was the short lamp with the Ikea reflector

718 910 704
770 1140 780
667 824 669

Average = 798 LUX wow my heat glass really kills more than heat!

The short lamp is the LUX king with the reflector it gets 594 LUX the best the long lamp got is 522 LUX

Removing the heat glass made the greatest difference of all.


The results for the short lamp could be even better if you adjusted the lamps for lumen output.

My helper taking LUX readings.

Click to view attachment

Now for "narrow arc vs. fat arc".... Good Job AV!

Excellent, excellent, excellent, excellent!!!!!

Great work. It shows two things. The first is that the pj DOES focus on a much larger portion of the lamp than just a small part. The second is that there are some benefits to a shorter arc, most specifically from the greater % of the arc useful in the reflector (seems the reflector is more picky about arclength than the fresnal).

Let's analyze the situation.

your long arc puts out 13,250 lumens. Your short arc puts out 12,000 lumens. On a total lumen basis your long arc bulb puts out 110% of the lumens your short arc bulb does.

When you look at what's being tranmitted without the reflector you get 401 Vs 380, about 106% more from the long arc bulb. the conclusion is that the fresnal may have missed projected the extreme ends of the long arc bulb. But this does say that the fresnal caught most of the 60mm filiment. This means that overall lumens is more important than arclength, as long as the triplet can capture most of your arclength. the results of the reflector basically show us that the limitations of the reflector must also be accounted for. But the most important conclusion from this is that we now have an answer to our age old question: Is it total output or lumens per length that matter?

So, let's see if I can put together a quick equation for this. The amout of useful light to start out with is: (arclength that fits in triplet)x(total lumens/total arclength)

This should start yielding much more accurate results.

Great work!

What did you use as your collimating fresnal?

The Fresnel was 650mm front 330mm rear. It also means that the more efficient the reflector the greater the hit you will take for the long arc.
I saw a long toy drop light that had the coolest oval shaped reflector but I think it was plastic It might have been good for longer arc lamps but my wife was looking at me kinda strange

Also the longer the rear fresnel the longer the arc and the wider the reflector can be.

Arizona -

You indicated that the bulbs were 13,250 and 12,000, which are the values given at bulconnection.com but did you test those values without any optics just the bulb radiating into free space to see if the bulbs actually meet thier rated specs? I'd put the lux meter at 1 foot or 1 meter whatever is convient and check to see that they line up with the specs.

As yo know I think that short arcs are the answer but these results seem to show that while the short arc won it didn't win by as much as I would have though especially given a 60mm filament. These results do seem to suggest that a short arc is brighter for similar power but the long arc isn't far behind and so a 600W or 1000W would still see a good amount more light than a 400W.

well...a couple of things in relation to the 1000 watt question... notice that the arc magnification is 2:1 (650/330) so the 60mm arc is a virtual 120mm arc and fits in the 128mm triplet entirely, if he'd used an 80mm arc he would have lost some percentage from the arc not fitting in, AV , you might want to try a 80mm arc and see what percentage is lost, might help solve that issue ..also unless my math is wrong, the ikea was twice as efficient(?) with the short arc as with the long arc (380 to 594 short arc - 56% gain as opposed to long arc(401 to 522 a 25% gain, about what the LL pj gains fomr the ikea)...thats the part I find interesting...

edit: well, I was just thinking (too early in the morning for that ...it was with the same reflector for both...perhaps with a bigger reflector the longer arc lamp would do better in percentage gain? maybe the arclength to reflector ratio has some optimum value, seem slike it probably does...I think the ikea is about 100mm or so so the short arc = 13/100 13% and the long arc = 60/100 = 60%....

pagercam: I think thats what he did with the measurement at 3 feet (unless I read that wrong) at 3 feet the area of a sphere is about 10.5 square meters so his numbers were about 50% higher (roughly) than the stated output, which oddly enough is the same percentage my lux readings were higher on a 60 watt bulb were due to room reflections.... maybe we've discovered a new "constant" the "room reflecto ratio"

Sorry I missed that part.

Also, a great finding here is that the heat shield reflects 20% of the incident light, not 10% as most of us assume. A redesign of the cooling system to eliminate the heat shield should net a 20% gain in lumen output. So, an extra fan or 2 for 20% more. Not a bad tradeoff.

I'll say it again. Great Work AV!

what kind of heat glass were you using AV? lexan or ?

Interesting find. I should try pulling my Lexan and see what that yields in terms of lux at the screen. (Don't want to run it that way for long, tho.)

Let us know what you get Supra.

I just pulled my Lexan and WOW what a difference.

My Luxmeter is broke (stepped on the white cover and broke it off. But the readings I get without the cover at the center of the screen with Lexan was 81 , Now without Lexan it is 93. Which indicates a HUGE difference.

Now my question. I've been running the projector without the Lexan for 20-30 minutes . Temp readings are 105 degrees at the rear fresnel on lamp side. 88 degrees between both Fresnels. And 78 degrees at LCD. Is this fine ? Will the rear fresnel warp at 105 degrees ?

Oh and the brightness increase is quite noticeable from memory comparisons of dark scenes I'm familiar with on several movies.

I wouldn't think so. Most people report around 95-97 degrees. But the only way to find out is too watch a movie and see if the you start to get hot spots after a while. If not, your golden.

Light transmission of XL10 Lexan is 88%.

Sorry I meant rear fresnel.

You must use a UV filter to protect your LCD panel from premature aging.

Has that ever been proven ?

I read back on a lot of threads and it was always conjecture but no proof. And the conclusion was that nobody knew.

Or am I wrong ?

Well, I was under the impression that some of the lamps have UV protection in the jacket, some don't.

For those that don't, what about using something like this:

LCD Protective Sheet

Could go on the LCD itself, or possibly on the smooth face of the collumating fresnel that points towards the lamp. It can handle the heat, and they claim 99% transmission of light. Interesting...

anti-reflective finish = lower light output

My question is, are IR blockers more efficient if the light is relatively collimated? As in, would it be worth more light to put an IR filter between the fresnel and the LCD? I know it would shorten the life of the fresnel, but really, they're not that expensive if you needed to replace one.

Can't exactly put the heat shield there and expect everything to be ok, but seems like it would be fine for an IR filter.

This is my heat glass It is for a 10" 2000 watt Colortran fresnel spotlight It realy stops the heat 2 or 3 times better than lexan but you can see the split so I may not be able to use it in the final PJ Hear is a shot next to my new lamp

Click to view attachment

The main thing I want to to here is redo the lamp LUX and adjust for stray light. But I am going to post the new lamp and some first readings on the super thread.

Well it's been running over 6 hours and no warping whatsoever. And the increase in light is really nice.

I guess I'll be the guinea pig and let it run this way for a while.

What kind of toy was it. Cover that sucker with aluminum foil and see what it does.

It was a John deer drop light flash light vary vary shinny. maybe the next time I get to the train store.

seeing as how the long arc lamp gave a better reading when no reflector was involved, is it possible that a reflector that covered the full arc of the long arc lamp would have given a better overall reading?

Yes, Later on I bought a 8" Colortran reflector and I am still using it in my 1200 watt PJ mainly because the FL is long so it allows more space for the air to cool the heat from the lamp.

The larger the reflector the less the size of the arc matters and it will perform better. Longer arc lamps almost always make more lumen's too but this will be offset by better arc collection of a condenser lens system.

The problem with most long arc lamps is the low CRI and the fact that they limit the condenser lens you can use because the arc image will not fit in the triplet. A supper large triplet would fix this but we don't have one.

The smaller the LCD panel the larger (shorter FL) the condenser lens can be and the shorter the arc need to be.

A 17" LCD stops you from using a condenser much shorter than about 7" (If I remember correctly) So a lot of slightly long arc lamps will work. I would still like to see a 20mm 500 or 600 watt lamp. with a 90 CRI.


I have not seen this thread for soooo long! Thanks for reading it.

Thank a lot Arizonavideo for all the info.

I have 3 questions if you have a minute.

1- I am confused with your conclusion, you have better result (higher average lumen output) with shorter arc but you still work with longer arc lamp?

2- Since you are the lamp master … I am using a LL65k T15 lamp and I have connected it on a 400 hps ballast and it is working… do you think it is OK?

3- In my projector I am using a pro reflector, a 220 rear, a 17’’ tft, a 550 front and a 18” bee triplet.
I also have a 330 fresnel and a precondensor (not shore what FL).
Do you think I can get better out put using those two components?

Thank again.
Michelspascie

Most of the shorter arc lamps are just higher performance. They have higher CRI (better color) smaller arc's so they work better with condenser systems but they all have short life of around 1000 hours and low lumen efficiently compared to the ceramics so there is a choice to be made.

Some people have their PJ on all day like a TV. A 1000 hour lamp life would get used up fast. I only watch movies a few nights a week so a 1000 hour lamp will last me around two years.

I still like some of the longer lamps for people with the large panels like the 17"

The ideal setup for a 17" LCD is a 330mm rear fresnel with a 6x7 or 5x6 condenser and a lamp with around a 25mm arc or shorter.

With a condenser system you could run a 5x6 condenser and a 25mm arc lamp and the arc image will still fit in the triplet (The condenser and fresnels make the lamp arc larger at the triplet so you have to make sure that most of the light will fit through.). At this point I think a 250 watt HQI with a 5x6 condenser might be brighter than a 400 watt condenser system mainly because if you you had a shorter fl high heat condenser for the 400 watt lamp the arc image from the lamp would still over flow the triplet so their would be some loss.

A 400 watt lamp system needs a high heat Pyrex condenser($40 to$50) so it won't crack. I have not seen a 7" fl one yet only 9" which is a little too long of a FL so you loose light.

In a dark room a low LUX PJ might be OK but it takes some work to get higher LUX. A condenser and a higher performance lamp like a 575 watt HMI lamp will always work good but there is some that are happy with the standard 400 watt and a condenser and get thousand of hours of use.


For a 17" my first choice right now would be, Pro reflector, 6000K HQI 250 watt overdriven to 300 to 350 watts, 5.5x7 condenser ( I happen to have one), 330mm rear fresnel and any front fresnel and triplet for your throw.

Second is the 575 watt HMI and a high heat condenser. This cost more and makes lots more heat and has short lamp life but the lamps make great color and the picture will be nice and crisp.

Third would be one of the 400 watt ceramic lamps and a condenser lens. there is a EYE ceramic that has not been tried and a retro lamp that makes nice red.

If you only watch 16/9 ( you can lock you screen in windows) you can cheat and use a 220mm rear fresnel. A 220mm fresnel and a 6x9 condenser will work and be nice and bright but for a 4/3 screen the corners will be black. You will need more cooling for a 220mm setup.

Their is a 250 watt ceramic that has not been tried either.

http://www.eyelighting.com/tb/MH/CeraArc/E...52-78-55108.pdf

I wish the ceramic had a higher color temp but 4000K will have to do.(only slight overdriving)

So far I have been called a Lumen legend and lamp master, which is nice I think.

EDIT.

Running a T15 on a HPS ballast might make over 400 watts but I have not done it. If the lamp is not slightly green or blue then I would guess that it is at full power or maybe slightly more. This is a fairly large mismatch but is better than the other way around. I would call it OK and perhaps a easy way to make a free extra lumen.

Any condenser lens will give lots more light but non-tempered lens often crack in a 400 watt system and will have to have a fan directly on them on the hot side to cool them. If the lens has a longer than 10" FL it might work with a 220mm fresnel. Test it for fl by focusing it on something bright like a lamp and place the lens about 10' from a wall and move it until you see the lamp across the room in focus. It helps to only use the middle of the lens so you can mask off the outside half if you want.

If the fl is shorter than about 10" you will have to use a 330mm rear fresnel.

Dazzla has a nice program to find the right condenser for any setup.

http://www.lumenlab.com/forums/index.php?showtopic=11256

I like the 6x9(FL) condensers from the 6" elliptical spotlights if you can find them. Most are high heat.