Trying to verify Maukka's sphere was accurately calibrated...

I have edited this and was going to get it deleted.

I went back to Maukka’s thread showing the test cert for his calibration light he used to calibrate his sphere, BUT who knows whether that calibration light was accurate being from China. Not to say something from China is an issue, but it’s an interesting thought when most of us with homemade setups are using a maukka tested light.

Can we really say that a manufacturer is wrong in their specs, when we don’t have a professional setup with a verifiable calibration light in hand?. Imagine if we’re all wrong… I know DB Custom was unsure about the accuracy when he first used a Maukka light since it was a much lower number. If we see 2000 lumens for a 2000 lumen model, this don’t mean our setup is accurate. I know some brands rate their lights lower, perhaps to account for any lower bins used in future batches.

And what have you found out? Does he have proper calibration for his sphere?

I went back to Maukka’s thread and found his image of the calibration cert. Still remains whether that Chinese calibration light was accurate though lol. I’ll never know and doubt anyone here can help verify that

You could send your Maukka lights (or something else) to a professional lab to get a full measurement on your lights.

You raise some good points there Funtastic!!

That is a great idea. Now to find a lab in New Zealand. I’d happily pay to know if it’s accurate.

On my TA sphere, calibrated to my S2+ Maukka light and using the stock meter, the TKLamp calibration source shows as .918 what the calibration sheet says it should be. It needs to be multiplied by 1.0893 to show the correct number.

Going forward I’ll be using an Extech EA30 meter with my sphere and the TKLamp calibration source to calibrate. Being a 2100 lumen source, it offers ~8x the resolution of the Maukka lights.

I also use a Maukka light.
My sample seems to be fairly accurate since I also calibrated my sphere with several LEDs of different flux bins, manufacturers and batches and the results are matching fine (± 3%)

My TA tube has been fairly accurate when comparing values to spheres on 1Lumen, a little less but not by much. The first batch of these tubes did have issues and new diffusers had to be shipped out. I received one of the last ones.

That TKLamp Sphere, have you compared numbers to another reviewer with an S2+ Maukka light and Sphere? I wonder how accurate it is.

My opinion is that the Maukka calibration lights were simply not well done. Better sources should have been selected.

After charging the battery in the BLF348, the output numbers increased by about 50%, so it is no longer in calibration. This is contrary to the claim in the original sales thread that it stays within 1% when different cells are used at full charge.

The Convoy S2+ works better as a source, but the values I get for 0 seconds, 30 seconds, and 60 seconds do not change at the same rate as on the included measurement report, nor does the total change correspond to the report. In short, this is not a reliable calibration source.

Further, an output value of ~270 lumens is too low to calibrate a sphere for high output lights. If the high mode were also tested, that would have been an improvement, but there is not enough thermal mass in an S2+. A copper host, or a solid mass of copper, or a fan cooled heatsink would be better. Avoiding PWM would have removed another possible variable that could affect results between different lux meters.

1 Thank

Yeah, but I think it’s okay for what most reviewers bought them for. I personally just wanted to test lumens for the Convoy flashlights I sell in my store. Simon never tests lumens and just states X lumens for X current. As long as it’s within 10% I’m quite happy.

Might get a TKLamp, but in NZD it’s quite expensive. Close to $1000 including shipping.

Edit: I see they don’t ship out of the USA anyway, so there goes that idea.

I do not have the TKLamp sphere, I have their calibration source and am now using that to calibrate my lumen tube.

1 Thank

That looks like a good idea as a calibration source.
Being driven off of 110/220 voltage as opposed to batteries is huge upon itself.

We can go round and round about what is the “True output/accuracy of anything” but at the end of the day, all these devices are a Guide and not a Bible.
They provide a Baseline, nothing more nothing less.

From the description of that calibration device;
Luminous Flux: 2100 ± 100 lumens
*Input Voltage: 100-240V *

Calibration lamp is calibrated in commercial sphere in TKlamp lab. NOT accredited calibration.

So, here allowable variation is 100 lumens +/- which to me is reasonable but it is still a swing of 200 lumens total from spec.
If you buy a Dodge Hellcat and it is rated at 807 hp Factory, the chance of you getting that Exact reading at you local speed shop Dyno are very,very slim.
Could be over, but most likely under. Does that mean Dodge lied? No, not at all but it may be the best reading they had under ideal conditions when tested.
Same thing here with all these little things we Call Flashlights.

Mass produced items subject to normal manufacturing tolerances.
And so, our relatively low cost Lumen meters and the Calibration lights recommended are expected to be the same deal.
I like my Maukka’s Lumen Sphere, just wish I use it more :slight_smile:
Have one of the last ones out the door with the updated diffusers.
Do not use it for anything professional it is just another tool to see if a Mod made a significant difference, etc.

The nominal output of these devices is 2100 lumens, +/- 100 lumens. This does not mean that the output of any particular source fluctuates by +/- 100 lumens. Each source includes a calibration report with an exact output value, which on mine says 2075 lumens. The output only varies by a few lumens during the warm up routine and is consistent from run to run.

1 Thank

That is fantastic!
Great way to test the tube then.

How much does your tube vary in different humidity and temps? Just curious.

What kind of light source is in there? Incandescent lamp, halogen lamp, LED? Are there any additional (driver) electronics, apart from the power supply unit?

COB LED, mounted to a fan cooled heatsink. The power supply is a Meanwell CC LED driver, model APC-25-500. I have not taken the lamp housing apart to see if there are additional electronics in there for the fan.

I inquired to TKLamp about the exact temperature of the heatsink during the calibration, and where I could attach a probe to verify. They only said that ambient should be 25C and warmup time should be 15 minutes.

It shrinks in cold water.

1 Thank

LOL
Good One!

Ok, an LED is not the best choice for a calibration light, simply because of the changing light flux with rising Tj.
The fan does not really help either since the light flux is still dependent on ambient temp and fan quality (if it is replaced by another fan due to defect or something the calibration is not valid anymore since the heatsink could have another temperature as in calibration process and therefore different light flux - difference between 2040 or 2120 lm for example is still a difference at all, which can have impact on test results in this sphere).

There is a reason why still incandescent or halogen bulbs are used for reference sources / calibration lights on most commercial lab-grade spheres.

1 Thank