First Impressions 2019 RRT-01 w Red 620nm XP-E2

2019 RRT-01 mod to Red LED 620nm
I reflowed to the stock mcpcb… no drama.
The LEDs came from azhu, same as my sw45k

with stock XP-L LED on 16340
500 lumens on my meter (NOT the 950 spec)

with Red LED
250 lumens on my meter

In actual use the red beam is Very Throwy, I can easily see a wall 15 yards away in the dark using only 10 lumens Red.

I can also see the wall using 10 lumens from my sw45k in 2012 RRT-01 w moderator007 reflector, but it is dimmer, and the hotspot is larger.

and when I measure the hotspot on a wall up close, with both lights set to 10 lumens, the hotspot from the red LED is almost twice as bright as the hotspot from the sw45k

the red LED has a Much more focused hotspot.
sw45k on left:


620nm Red:

I like the Red! The color is rich and I can easily see with it.
Im happy to report, nothing melted during my brief tests of maximum mode.

Thumbs UP!

I like it, nice mod. :+1:
Which driver did you go with using the red xpe?
Just wondering if you noticed the same thing I did with the modded red led Ti3, the red seems way brighter than what my lumen tube reports.
I dont know if its the Extech LT45 isn’t picking up the 645nm red correctly or if its just my eyes not being strained to see with the preserved night vision or if the meters just reading correctly and I’m terrible at perceiving lumens. In high mode my lumen tube reported 30 lumens but it looked way brighter.

your red xp-g3 mod is cool too :slight_smile:
I kept the stock 2019 driver… nothing melted

I agree the red looks a LOT brighter than my meter says… like double, I mean that literally.

The red hotspot is twice as bright on the wall as my sw45k hotspot, even though both read 10 lumens on my meter, the red just “looks” more intense.

so I dont know if the meter cant see the red as well, or if you and I are reacting to a more focused hotspot having more LUX

but whatever the reasons… Im totally impressed that this little red led can handle the power, is WAY brighter on max than I need, and im a Big fan of the tiny hotspot.

this is only the second 2019 RRT-01 Ive handled. Given that the Original are sold out, I think this is a perfectly serviceable little Rotary.

when it arrived, the control ring had issues… the ball had come off the spring inside the control ring, and the dial would stick half way… I managed to free it up before disassembly, and found the ball bearing was on top of one of the magnets when I lifted the control ring.

So I dont know if the initial stock xpl reading of 550 lumens on 18350 was due to the ball interfering with the magnet. This light was supposed to make 950 lumens stock.

In any case, I can tell you, and anyone that cares to listen, that the Red LED at 250 lumens is ridiculously bright, I would never use it at that level. But I did, just to see if I could melt the LED… so far, no issue.

I find in practice that 10 lumens of Red is plenty to see everything within 10 yards, easily.

Well, not everything
Red things kind of disappear

That Huichol beaded bowl I posted above, with the red, I could not see any of the orange beads in the middle…

but, Red is not about seeing colors, its about preserving Rhodopsin.

I can certainly navigate walking around with red light. Its a bit different though, some red things disappear.

look at these beads under my house lights… then watch what happens to the orange and yellow beads

when I add the red light:

they totally disappear!

Jon using a LED with –75 R9 value? Now he has the modding bug for sure

Luxmeters up to pretty expensive ones make large errors in the blue and red regions of the visible spectrum (in the order of many times over- or underreading the actual luxvalue, you need a DIN classA luxmeter to get any accuracy). It may even be that visual estimation of brightness is more accurate than your luxmeter reading.

If you stick to one wavelength though, luxmeters have their use to relatively measure brightness because while their absolute value can be very wrong, they are very lineair in their reading.

I swear its not my light… LOL
I only did it for a friend, honest :slight_smile:

but yes, Im having a really fun time swapping LEDs :slight_smile:

well, that sounds true
now, riddle me this

both these lights are set to 10 lumens out the front, when each light is sitting on my meter

but when I measure the brightness by putting the meter at the wall in the middle of each hotspot, the Red reads 2.1 lumens, that is 62% brighter than the sw45k that reads 1.3 lumens…

so… what do you think is going on?
my guess is it has everything to do with the hotspot being more focused on the Red, so as a beam, the center is brighter, but in overall total out the front, the red is not reading more on the meter…

weird huh?:wink:

now these photos do not have matched lumen out the front, I just made them visually similar… both lights are actually turned down to very dim outputs,
note the red LED is smaller:

in this next pic, I did not match the outputs on my meter, just wanted to compare colors, the sw45k in the TiC on the right, is with the moderator007 reflector upgrade (that fits Original RRT-01, and Eye10):

Sounds indeed what is going on.

The XP-E2 die is twice as small (surface area) as the 219B die, so the hotspot should be smaller, as you found.

Did you change the light source selection to red LED? (dont forget to change it back or else your calibrated one will be wrong!)

I had thought this was a decent mid-range meter and should read the red wavelength. I just stuck it in the lumen tube and took a measure. I had forgot all about the LS seletion. I’ll change the settings and try reading it again but I’m assuming without a calibration red light that number will still be inaccurate.

cant you calculate it since you have the stock white settings, calibrated lumen setting and stock red setting?

I was curious about running this led with the 2019 driver until I came across djozz wonderful red xp-e2 test.
Apparently the red xp-e2 has a steep vf slope compared to the current.

So I’m guessing that’s the reason it worked without melting anything. Looking at the graph suggest this led will work great with this driver or almost any driver that has a output of 4.2v or less. Even direct drive looks like it would work from a single cell, even though it has such a low rated vf. I would have never guessed that without djozz testing data. Thanks for the test djozz :+1: .

Could you elaborate a little more. :person_facepalming:

L0 0.958 incandescent
L1 0.851 LED white
L2 0.902 LED amber/yellow
L3 0.772 LED green
L4 1.642 LED red
L5 1.475 LED blue
L6 1.148 "Not used"
L7 0.484 my custom bobbersphere setting

I have the LT45 too and when I look at my settings I get the above. L7 is set to my own factor based on a calibration light and diffusion I used in my bobbersphere.

If a red LED is approx 1.93x a normal daylight white LED (0.851 * 1.93 = 1.642) cant we multiply our integrated values by 1.93 to get "red lumens"?

I suspect you could gut check this with the datasheet of a red LED.

Also here is the meter sensitivity by wavelength:

I guess if we say LED peak is around 580-610 and red LED is 620-660 it seems reasonable? I'd bet the 1.93 is more for standard red and photo red would be even higher than 2x to be honest.

That Extech graph is very coarse and not suitable for extracting calibration information. But assuming that different Extech LT45 meters have the exact same error, my luxmeter test of the LT45 gives some insight, here are two graphs from the tests. First my Mobilux meter versus the defined Vlambda curve:

If we trust the measured data that I got from Czibula & Grundmann (manufacturer of the Mobilux measuring head), the optical filter/sensor combination of the Mobilux luxmeter is nearly perfect for all wavelengths so I can use it as reference.

Now the LT45 measurements of a Tungsten lamp together with the Mobilux measurements of the same Tungsten lamp, the surface areas under the graphs were normalised, so a white source with a flat spectrum would measure the same in all meters (this introduces some systematic error since they are calibrated for a 2700K tungsten lamp which has a sloping spectrum):

You can now use the relative difference between the tungsten spectra of the two meters for a calibration of any wavelength, i.e. the LT45 measures twice too low at 660nm and 40% too low at 625nm.

There are some uncertainties, like does my Mobilux have the exact same filter/sensor precision as the measured one by the manufacturer where I got my data from (I sure hope so for its price :neutral_face: ), same for different LT45’s, and is the normalisation of the curves correct enough to derive a calibration this way.

But my feeling is that it is a fair estimate of what is really going on.

I see now that the L4 setting for red leds quite agrees with my measurements, the L5 for blue reasonably agrees for 470nm (but not for 450nm) but for green it is not so close.

Who says you can’t teach a old dog new tricks. :smiley:
I learned some new things today, thanks djozz and contactcr for the well detailed info. :+1:

Sorry for getting side track, jon_slider.
Hopefully we didn’t completely jump the rail :person_facepalming: .

Im happy to watch and learn
I agree the red was very bright

it makes sense to hear my meter was reading low on Red

I think the dimmable red strobe in the 2019 RRT-01 is fun.

now I want a green one :person_facepalming:

Maybe an Osram green one for extreme throw? I don’t know if it will be able to handle the amps though. Maybe with the older driver…

I could see using an AAA light, no issues with excess power that way. But since we know the xp-e2 red can handle the 2019 RRT-01 power… I would go there again if they make a green (have not learned to locate them yet…)

fwiw, here is a thread about the benefits of Red Light…