Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

[wbp]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

If I used the exact sphere as you and picked up a light that is consistent and used the same meter. Couldn't I put the light and meter in the box then read the number then ask you what the "right" reading should be (since you calibrated it). Then I assume in the back of the meter, I could adjust it?

Sorry if I misunderstood.

No two spheres are going to be *exactly* the same; even a pro IS is sold with calibration or some provision for calibrating it in the field. Ask Gman how often his company's IS goes back for re-cal.

To measure lumens you don't need to adjust the meter. All you need is the correction factor for your meter and sphere system to put in your spread sheet.

Even with the same brand and model of light there is still variation in output due to the LED variation. If you want to measure lux accurately you have to have the meter calibrated. I can do this but it takes time and expertise I have had to develop, and often components have to be changed as well.

Don't forget - I've got $$$$$ invested in test equipment that allows me to do this. I want to help you guys out as much as I can, but I can't afford to give it away either. I would like to (need to) recover some of my costs... I think if enough people are interested in doing this then we can come up with something that is both reasonably accurate and reasonably affordable.

 

[MrGman]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

Okay I am in the mood to test more lights. I haven't seen the local guys with the Surefire E1b whatever models since last year so for all the E1B fans out there who want one tested, some one private message me for shipping directions and instructions if you want them tested. If there are different versions send them.

Try not to use to much post space just going back and forth on what to send, PM me.

Other Surefire lights are welcome. I did test and post the "CPF control A2" with a steady 65 lumens.

Any of the smaller 1 CR123A or 1AA that is not a complicated programmable light would be fine. But I know a lot of guys have been wanting the E1B tested, so bring it. G.

 

[SilverFox]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

Hello Gary,

I was unable to find the value for the SureFire A2 LED's. What color are they and did you measure them?

Tom

 

[MrGman]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

Hello Gary,

I was unable to find the value for the SureFire A2 LED's. What color are they and did you measure them?

Tom

The A2 I measured was the center incandescent unit only. It was the one you sent(?). No LEDs were installed, they had been removed. Did I call it by the wrong name?

 

[SilverFox]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

Hello Gary,

You should note that when the LED's are removed, the lumen value goes up. Simply state that it is the CPF Benchmarking A2 and that should cover it. It is not representative of the off the shelf A2 because it has the LED's removed.

Tom

 

[cy]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

just found this thread... been awhile since I've cruised new posts...
nice work!

 

[MrGman]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

Just a quick update. wbp and I both now have 16 inch inner diameter stryofoam (polystyrene) spheres. I also have the same CA813 portable meter he has (yet to be cal'ed to his). We can now both take in 3 inch diameter head LED light sources to do testing for total lumens. We both have several calibrated flashlights I have tested in the Integration Sphere system at work. So now the fun begins in looking at all the larger lights and see how they track to the smaller ones and to each other all measured in the same sphere system.

wbp's testing has shown that of all the home made materials one would want to try including some highly reflective type paints, the polystyrene sphere as is works best. Stay tuned for future updates. G

 

[Superdave]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

I'll have to pick up one of these styrofoam spheres. My setup met with an unfortunate accident at work and got tossed.


Any pics of the new rig?

 

[MrGman]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

I have posted lumens numbers for a SureFire E1B for all the E1B fans out there. Its a F@CE model if that means anything to anyone. At least thats what the markings look like on the side of the head. Pretty handy light.

I have also been publishing the power draw numbers for the Malkoff MCE- warm and cool versions, Malkoff Triple Drop and M30 and M60 for you guys who want to try and run them with different battery configurations.

I will at some future point in time try and put the power draw and the lumens outputs together. I don't have a big enough DC power supply at home to supply the high current at low volts that I need just yet.

However I will say this with my new home made Integration Sphere which appears to be working very well. the Malkoff Triple drop is only in the 680 lumen range and its consuming very close to 15 watts of power (with some loss at the regulator) to do it.

The Solarforce L900M is in the 670 Lumen range and its drawing somewhere in the 14 to 15 watt range to do it. Mine draws 1.9A from 2X18650 to reach full power.

Eventually I will add a posting of all my readings from the 16" homemade integration sphere system. Still needs more fine tuning and testing.

Haven't had any time to take pics of it yet. Nothing special to look at. From the outside its just an 18 inch white styrofoam ball with a 3 inch hole on the side and the 1 inch hole on the top covered with the light sensor from the AEMC CA813 meter. I left it sitting in the cardboard box it came in for support and simply cut a 4 inch hole in that to allow access to the entry port. I am using several of the lights I measured at work on the real IS system as my calibration control units.

Just as another point of refence we took bigchelis' "Nailbender P7" that was actually measured in my new 16" sphere system at 456 lumens (meausred 460 in the real IS) and drove it with a megabattery pack through a series of resistors. I lowered the resistors in series until I got it up to 4.75V across the LED module. Output jumped up to what we calculated to be in the 689 lumen range. Current draw was not measured but I know from previous measurements it had to be over 3 amps. Point is in order to get that P7 unit up into the same type of "through the glass, out the front" output range of the other lights in that output class, it also had to be in the 15 watt power consumption range. And it was getting very hot very quickly even inside the flashlight host with only the rear tailcap removed to allow the jumper wires inside.

These discount P7 or MC-E lights that fit into a normal 6P type host and run off of 1X18650 class battery are never going to reach 700 OTF lumens because they don't have the drive power to do it and don't have the heat sinking to sustain it if they did. The ones with massive heatsinks, 2X18650 or more will begin to reach that as OTF lumens, with the die actually hitting 900 lumens or more and the typical losses through reflector and glass.

As always more to come at a later date.

 

[wbp]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

I'll have to pick up one of these styrofoam spheres. My setup met with an unfortunate accident at work and got tossed.
Any pics of the new rig?

Here's where to get the 18" (OD) sphere we are now using:
http://www.barnardltd.com/product.jsp?prodId=1878&catId=

Pictures coming soon! (we hope)

 

[polkiuj]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

MrG, what about E1B driven with a fresh RCR123 (16340)? :naughty:

 

[MrGman]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

MrG, what about E1B driven with a fresh RCR123 (16340)? :naughty:

This model belongs to some one else who is not willing to risk it and I am not going to risk some one elses light for them when they said no. If you have one that you do it to all the time and want to send it send me a PM.

 

[MrGman]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

http://www.flickr.com/photos/20634664@N02/sets/72157617408605639/show/

modest set of Integration sphere pics are up. Like I said not much to look at from the outside. You can see how well it lights up on the inside just around the port which is only 3 inches diameter.

 

[TexLite]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

http://www.flickr.com/photos/20634664@N02/sets/72157617408605639/show/

modest set of Integration sphere pics are up. Like I said not much to look at from the outside. You can see how well it lights up on the inside just around the port which is only 3 inches diameter.

Wow this is great. Thanks for the pics, very nice. Are you planning on starting a new thread or is it ok to continue discussion about the setup you guy's are working on here in this thread?

I'll limit my comments/questions in case you want to move the discussion...

The sphere looks great, do you guys think the 18" is the way to go?

Could you post some info about the baffle placement in relation to the access and meter port?

I know you guys are using the CA813, but the reason I chose the Extech 407026 is the PC interface, which is great for runtime graphs. The specs look similar between the two meters, and I'd be glad to send it to you MrG or wbp if you guys wanted to test it out. The reason I mention this is to make a suggestion. If anyone wanted to go with the Extech, don't buy the datalogging software and cable from the manu./dist. The cost is an additional $80.00 which could be bypassed for pennies on the dollar. The cable can be easily made up for around $10.00 with RadioShack parts and the software can be downloaded free from Extech! Had I known this before I could have saved some coin, so I just thought I'd pass it on

Thanks,
Michael

 

[MrGman]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

I don't have the baffle in mine. Wbp has it in his half way between the 2 ports. He used the piece of foam cut out from the 3 inch port reshaped to fit the inside of the spheres curve on the bottom portion and simply stuck on with doublestick tape. We are comparing his readings of various lights to mine to see if it makes a difference. Since the units are only held together with masking tape, opening them up again to make changes is no problem. The sensor does not come all the way through the sphere wall because of the wall thickness and light cannot directly hit the sensor where it is so it kind of already has a baffle. We don't know if this is enough, but since the spheres are the same and the type of meter is the same, running all the same lights I have tested in mine versus his should give us a really good indicator as to whether or not the baffle is necessary in a sphere of this inner diameter. We will post those results soon.

I will discuss with wbp about cross referencing to your extech meter. Send me a PM as to exactly what model it is, does it have cosine adapter and how wide is the sensor on it. My sensor port is 1 inch and I am not going to cut that wider for a different sensor type as that will change everything for my sensor. G

 

[kramer5150]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

Just as another point of refence we took bigchelis' "Nailbender P7" that was actually measured in my new 16" sphere system at 456 lumens (meausred 460 in the real IS) and drove it with a megabattery pack through a series of resistors. I lowered the resistors in series until I got it up to 4.75V across the LED module. Output jumped up to what we calculated to be in the 689 lumen range. Current draw was not measured but I know from previous measurements it had to be over 3 amps. Point is in order to get that P7 unit up into the same type of "through the glass, out the front" output range of the other lights in that output class, it also had to be in the 15 watt power consumption range. And it was getting very hot very quickly even inside the flashlight host with only the rear tailcap removed to allow the jumper wires inside.

These discount P7 or MC-E lights that fit into a normal 6P type host and run off of 1X18650 class battery are never going to reach 700 OTF lumens because they don't have the drive power to do it and don't have the heat sinking to sustain it if they did. The ones with massive heatsinks, 2X18650 or more will begin to reach that as OTF lumens, with the die actually hitting 900 lumens or more and the typical losses through reflector and glass.

As always more to come at a later date.

I was thinking about modding a DX MC-E with a ~2.5A driver board. Is that a futile effort given the capabilities of 1x18650, and the thermal limitations of the SF-6P?

Fascinating thread.. thanks so much!!

 

[wbp]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

The sphere looks great, do you guys think the 18" is the way to go?

Could you post some info about the baffle placement in relation to the access and meter port?

I know you guys are using the CA813, but the reason I chose the Extech 407026 is the PC interface, which is great for runtime graphs.

Michael,
Yes, I think the 18" (16"ID) is the way to go. You could use the 12" (10"ID) but the 18" is not that much more money and should give better accuracy.

I will open up my sphere and post pictures so you can see the baffle. I agree with MrG that it might not be necessary due to the depth of the hole for the sensor, but have some concerns that direct light from the source being tested might hit the hole and bounce, causing errors. The baffle is trivial to add so why not? It does reduce the reading slightly but not enough to be of concern.

I would be tempted by the Extech if I didn't already have an SP-100, which can easily be automated. Does Extech provide an API or activeX component to help with programming?

All of these inexpensive meters are very poorly calibrated, and that's true even if you pay for NIST certificate! We've compared several and the readings are all over the place. I don't know how they manage to make the specification claims that they do and get away with it - I guess no one is interested in taking them to task (or court) over a $200 meter.

I am curious to see the Extech so if you can spare it for a few days let's discuss offline.

william

 

[polkiuj]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

Hey MrG,

We have an interesting "discussion" going on in another tread about the Fenix PD30. Mind testing it once and for all so we could all save our saliva (sweat and keyboards) and sleep better in the near future? =D

Thanks!
Walter

Disclaimer: Unfortunately, I don't have a PD30 and I live in Malaysia. =(

Note: I want to send my Fenix LD10 for testing as well (in the near futture, as I don't have any other lights now... yet). I fully believe it does nowhere near the claimed 120 lumens as my Inova T1 (it is being sent back for a warm tint) just kills it. But what I'm really interested in is running it with a 14500 (hehe, and I will provide it =D). What say you?

 

[sygyzy]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

I am very excited at these developments. Thank you for your plans to make it "open source."

Sorry wbp - I don't you want you to feel ripped off or that you're losing money on it. I am sure you two can come up with a solution that wuold be benefifcial the community and still allow you some income.

I am a real numbers guy and would love to get some system together to measure lumens (I got it now) with some reasonable accuracy.

 

[MrGman]

Re: Actual Lumens readings in 6" dia. Lab Sphere IS with SC 5500 control - PART II

Hey MrG,

We have an interesting "discussion" going on in another tread about the Fenix PD30. Mind testing it once and for all so we could all save our saliva (sweat and keyboards) and sleep better in the near future? =D

Thanks!
Walter

Disclaimer: Unfortunately, I don't have a PD30 and I live in Malaysia. =(

Note: I want to send my Fenix LD10 for testing as well (in the near futture, as I don't have any other lights now... yet). I fully believe it does nowhere near the claimed 120 lumens as my Inova T1 (it is being sent back for a warm tint) just kills it. But what I'm really interested in is running it with a 14500 (hehe, and I will provide it =D). What say you?

Some one in the US with the Fenix P3D or PD30 will have to step up to get it tested. I am not buying any more lights unless something really fantastic comes along.

You didn't mention the other thread where this discussion is going on.

I wont be shipping any lights back to Malaysia. If some one in the US has an LD10 and the special rechargeable batteries that wants to see if it runs brighter and by how much they can PM me. This type of light isn't even close to the list of what I would be interested for myself but I will test it because "it needed testing". G