Destructive Incan Bulb Tests - Updated 8/27/2010 (Newer Info Added)

Nite

Enlightened
Joined
May 27, 2008
Messages
845
Location
NYC and Long Island
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

Please PM me your shipping address, I would like to send a FM-1794 for you to destroy please.
lovecpf
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,661
Location
MS
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

Nite, I got those two 1794's today. Thanks! Man, those are tiny little suckers!! With some careful bending, I got them to fit in my standard KIU bulb holder. Just waiting to get those 3853 bulbs later this week and will do the tests.
 

Nite

Enlightened
Joined
May 27, 2008
Messages
845
Location
NYC and Long Island
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

Nite, I got those two 1794's today. Thanks! Man, those are tiny little suckers!! With some careful bending, I got them to fit in my standard KIU bulb holder. Just waiting to get those 3853 bulbs later this week and will do the tests.

I cant wait!!!

One of those has a few minutes use on it...it kept slipping out of the bi pin holder so many times I decided to have it murdered, by you!

Also, howd u like that packaging job? My GF works at a lab!lovecpf
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,661
Location
MS
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

I just did the FM CL-1794 because I know you are anxious for it, and I didn't get the Pelikan 3853 bulbs yet.

I see why everyone is flashing these on 2 x IMR/Emoli cells. These two bulbs had very tight correlation in measured Lux within 2 Lux at every voltage. Both flashed as soon as I went to 7.9V, so that is a measure of consistent reliable correlation.

The other thing I like to do is compare the default claimed Lumens and my measured Lux with other bulbs that claim around that same lumen level, and see what I get for Lux for those. When you do that, you start to see that a bulb like this is understating its default lumen, which all of my projections are based upon. The CL-1794 is listing default 628L where I measured 87 Lux. Compare that lumen level to other lux I measured to see what I mean. I estimate that the 7V output of this bulb is more like 1,000L

For example, look at the Lux I measured for the WA-1274 WA-1331 WA-1166 The Lux I measured correlates with higher lumen ratings all the way up the voltage scale.

For this reason--and this is the only time I have done this with these charts, I'm posting a second chart for this bulb of what I believe are the actual ratings. I don't have this light (regretfully), so the next step would be for some who have other 1,000-1500L lights to see how this compares to them on a practical basis. (One other time I lowered the manufacturer claimed lumen rating here)

I do think this makes it worth the $20 in bulbs that you sent me. You would think that FiveMega would want to contribute these, but oh well.

--------------------------------------------------------------------
Edit (5-23-09): I have since done my comparison shootout with other known Mag Mods, and lowered the estimated default lumens. However, this is still a VERY IMPRESSIVE performance by a very tiny bulb.





.
 
Last edited:

sami_voodoo

Newly Enlightened
Joined
Oct 31, 2008
Messages
43
Location
Nantes, France
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

Lux

Just a comment here, a quick calculation using the "new" lumen values that you calculated for the CL-1794 bulb give a rating of slightly above 53 Lumens/Watt (1333 Lumens and 24.96 Watts). I never thought these levels of efficiency or efficacy were possible for incandescent lamps. Am I missing something or are bulbs that are overdriven to such levels can give these outputs?

Cheers!
Sami
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,661
Location
MS
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

Lux

Just a comment here, a quick calculation using the "new" lumen values that you calculated for the CL-1794 bulb give a rating of slightly above 53 Lumens/Watt (1333 Lumens and 24.96 Watts). I never thought these levels of efficiency or efficacy were possible for incandescent lamps. Am I missing something or are bulbs that are overdriven to such levels can give these outputs?

Cheers!
Sami

It's a legit question, but remember my measurements should never be taken as absolute, rather they are relative to other bulbs tested in this setup. This result was striking (& yet consistent between the two bulbs) enough that I put an 1185 in the holder afterwards to make sure my LM-631 Meterman meter wasn't haywire, and I got very close readings to previous (posted) 1185 readings. I can't do any more now because the sun is coming up, and I should be in bed. More useful correlation would be to do practical side-by-side eyeball comparisons of lumens.

I forget what the Osram IRC bulbs can get up to when overdriven...I think it was over 60 L/W
 

JimmyM

Flashlight Enthusiast
Joined
Aug 30, 2006
Messages
2,853
Location
Boston, MA, USA
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

Someone needs to buy you an integrating sphere. :devil:
 

zehnmm

Enlightened
Joined
Jan 16, 2006
Messages
265
Location
DFW area
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

Someone needs to buy you an integrating sphere. :devil:

Agree. Does anyone know the choices of these devices, where to get them, and their price?

The last I heard they were over $5,000. While it could be a pipe-dream, if we could source one at a sensible price, I for one, am willing to pony up some funds and chip in. Should we get enough folks to chip in, maybe this could become a reality.
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,661
Location
MS
Re: Destructive Incan Bulb Tests - Updated 5/15/09 (Newer Info Added)

I think the cost of an I.S. of adequate size needed for all these bulbs with calibration checks, etc. is only justifiable for commercial bulb manufacturer, govt. researcher/contractor, or a multi-millionare who has satisfied their quest for the ultimate porn collection, and ready to move on to an obscure hobby. :laughing:

A good way to cross check my results is for someone who has one of these 1794 bulb setups and compare it to say a decent Mag85. If my readings are correct, it should hold its own against it, and outperform a Mag1331, Mag66, SF M6, etc.
 

KiwiMark

Flashlight Enthusiast
Joined
Oct 19, 2008
Messages
1,731
Location
Waikato, New Zealand
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

Just waiting to get those 3853 bulbs later this week and will do the tests.

The 3853-H bulb looks good from 2 x KD Li-ion protected D-Cells - nice colour, similar to my ROP (3854) low. I had been trying to decide what to do with my #3 Mag 2D, now I just have to decide what to do with my Mag 3D.

My ROP low is pulling around 2A and the 3853 High pulls around 3.5A - the brightness doesn't seem all that much different to my eyes though. Maybe a bit brighter but no where near twice the brightness. If the Lumen is proportional to the amp draw then we are looking at 75% more Lumen (by my eyes I would guess this is about right), but we need 300% more Lumen (4x the Lumen for twice the apparent brightness) to look twice as bright. Still - nice bright light with nice tint and over 1 hour run time on my setup, focuses well after I remembered to put the cam back on to the reflector.

Waiting to see the test results chart from Lux.:popcorn:
 

mdocod

Flashaholic
Joined
Nov 9, 2005
Messages
7,544
Location
COLORado spRINGs
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

The 3853-H bulb looks good from 2 x KD Li-ion protected D-Cells - nice colour, similar to my ROP (3854) low. I had been trying to decide what to do with my #3 Mag 2D, now I just have to decide what to do with my Mag 3D.

My ROP low is pulling around 2A and the 3853 High pulls around 3.5A - the brightness doesn't seem all that much different to my eyes though. Maybe a bit brighter but no where near twice the brightness. If the Lumen is proportional to the amp draw then we are looking at 75% more Lumen (by my eyes I would guess this is about right), but we need 300% more Lumen (4x the Lumen for twice the apparent brightness) to look twice as bright. Still - nice bright light with nice tint and over 1 hour run time on my setup, focuses well after I remembered to put the cam back on to the reflector.

Waiting to see the test results chart from Lux.:popcorn:

An overdriven 3854L would be probably close to the brightness of a regular driven 3853H.

Remember. Efficiency really cranks up as bulbs are overdriven.
 

sami_voodoo

Newly Enlightened
Joined
Oct 31, 2008
Messages
43
Location
Nantes, France
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

It's a legit question, but remember my measurements should never be taken as absolute, rather they are relative to other bulbs tested in this setup. This result was striking (& yet consistent between the two bulbs) enough that I put an 1185 in the holder afterwards to make sure my LM-631 Meterman meter wasn't haywire, and I got very close readings to previous (posted) 1185 readings. I can't do any more now because the sun is coming up, and I should be in bed. More useful correlation would be to do practical side-by-side eyeball comparisons of lumens.

The CL-1794 is listing default 628L where I measured 87 Lux. Compare that lumen level to other lux I measured to see what I mean. I estimate that the 7V output of this bulb is more like 1,000L

I get it. You're going by reference values and from lux values, you calculate the lumens.

Another question, I apologise if it's been already answered. Measuring an IRC-type axial filament bulb is simple enough, but what about bulbs that have transverse filaments? Do you take two measurements (at 90° each) and take an average? Do you even see a difference on the lux meter?

Cheers!
Sami
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,661
Location
MS
Re: Destructive Incan Bulb Tests - Updated 4/18/09 (New Info Added)

I get it. You're going by reference values and from lux values, you calculate the lumens.

Correct. While the Lux & Amps are actually measured at every voltage level, the first "Predicted Lumens" column is taken from the AWR Hotrater spreadsheet prediction
Predicted Lumens=Default Lumens *(Overdrive Voltage/Default Voltage)^3.5
AWR and I experimented with trying to find more accurate predictive calculations, such as this Excel cell formula:
=I9*(F11/F9)^(3+(0.55/(1+2.72^(5*(P11-0.5)/1.1))))
But ultimately, there is significant variation from bulb to bulb, and manufacturer to manufacturer, so that a "unified reliable predictive formula" was just not working.

It was at that point that it made more sense to go to a real measurement model, and this thread was born. I did as much as I could think of (short of using an I.S.) to control for variables. For example, that new Osram Ministar 50050 is not really a fair test from a side view because it has a built in reflector in the base.

Then I added the Yellow "Predicted Lumens from Measured Lux" which takes the % increase in my measured Lux from default, and multiplies that MEASURED increase x Manufacturer stated default Lumens value.

Another question, I apologise if it's been already answered. Measuring an IRC-type axial filament bulb is simple enough, but what about bulbs that have transverse filaments? Do you take two measurements (at 90° each) and take an average? Do you even see a difference on the lux meter?

Cheers!
Sami

I did think of this, and in some cases having the transverse filament perpendicular to the light meter tube gave a slight improvement. The whole goal here was to try and represent the bulb as a point source at 1 meter. It was not a significant change for smaller bulbs, but was notable with bigger watt, wider bulb filaments like the Osram 64623.

I did not want to do those additional parallel measurements because there was no easy way to keep distance reliable when rotating bulb holder mounted on board...so it would have had to be a whole separate set of measurements run for each bulb in fixed position with transverse filament perpendicular and again parallel to light meter tube.

I thought it more important to verify consistency between at least two bulbs, and doing additional filament orientations would have needed 3-4 bulbs for each chart. So wanting to give as much advantage as possible to the bulb, I just made sure to do all transverse filament bulbs perpendicular to tube.
 

sami_voodoo

Newly Enlightened
Joined
Oct 31, 2008
Messages
43
Location
Nantes, France
Re: Destructive Incan Bulb Tests - Updated 5/15/09 (Newer Info Added)

Thanks a bunch for your time! :thumbsup: Looking forward to the tests on the 3853.

Cheers!
Sami
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,661
Location
MS
Re: Destructive Incan Bulb Tests - Updated 5/15/09 (Newer Info Added)

OK, just got the Pelikan 3853 bulbs done and posted. A respectable showing with the High, but the Low is a girly-man bulb. You sneeze at it and it flashes.
 

ElectronGuru

Flashaholic
Joined
Aug 18, 2007
Messages
6,055
Location
Oregon
Re: Destructive Incan Bulb Tests - Updated 5/15/09 (Newer Info Added)

OK, just got the Pelikan 3853 bulbs done and posted. A respectable showing with the High

Great info. Looks like the High bulb has .2 volts greater range and about .6 less amps output overall, with correspondingly more runtime and less lumens.


the Low is a girly-man bulb. You sneeze at it and it flashes.

8.1!? :thinking:

This is confusing. I've poofed the 1111 bulb (rated to flash at 8.3) on the same cells/charge that both the Low bulbs handle just fine. To say nothing of the counter intuitive nature of the "better" bulb doing worse. Could it be an anomaly or bad batch?

I've got more 3853 on order (making 4 total tried). If all show just as strong, would you mind testing a proven one?
 
Last edited:

KiwiMark

Flashlight Enthusiast
Joined
Oct 19, 2008
Messages
1,731
Location
Waikato, New Zealand
Re: Destructive Incan Bulb Tests - Updated 5/15/09 (Newer Info Added)

OK, just got the Pelikan 3853 bulbs done and posted. A respectable showing with the High, but the Low is a girly-man bulb. You sneeze at it and it flashes.

I have tried a 3853-L on 2 x Li-ion D cells fresh off the charger - no prob. Maybe you got a bad one. I measured a drain of 1.6V. Seems to be a good bulb, fairly white, not much different to the 3854-L - a bit less amp drain for very similar output.
 
Top