Also remember than even 3 steps grouping is still an ellipse, no a point
In the data sheet, not all grouping ellipses follow the BBL symmetrically. The cooler temp slowly leaves the BBL as the temp raised. The 573 lower part ellipse barely touched the BBL. This risky balance is what makes up for better efficiency in 219C I guess.
For those who really donāt want even the slightest green sw45k 219B is the best bet, far away below the green zone. If you still want the the 219C, stick with the 3 steps and no higher than sm403.
May be, you know, when 219C (9050 3-step) sw40 & sw45 will be available ?
The SM403 4000k looks absolutely beautiful out of this C8-Triple!
The only Green I see is when itās OFF!
:+1:
Official information:
- NVSL219B-V1 Ra 9080 sw40 late April
- NVSL219B-V1 Ra9080 sw45k uncertain availability
- NVSx219CT ALL series late April
Thanks for the update. The Ra9080s are worth the wait! 
I think I better take shots of all the LEDās left from the last sale thread. Sadly I have no NVSL219B-V1 Ra9080 left to be included in the test. Later tonight, itās still too bright here
Thanks!.
Have a great day!.
OK guys, I decided to build a comparison light bar using all the left over LEDs from all my sales. It's a good thing for future reference. The LED installed are various CCT and CRI of 144AME, 319ATE, and 219CT only. I don't have any 219B left to test. But due to the lack of input voltage required to drive all of them (~50V total), I can't post all the LEDs tint shot for now. I need to built my boost rig setup later. Shot in various current from 0,01A all the way up to 2A.
I only post 0,01A, 0,1A, 0,3A, 0,7A, 1,4A, and 2A tint shots. Maukka was right, no noticeable tint shift to the naked eyes. I even confirmed these shots to some of my friends, they're all agreed.
White Balance set at my standard 5600K. Shot with Olympus OMD EM5, Zuiko EZ lens 12mm-50mm. Exposure was speed adjusted to give relatively equal brightness, no change in aperture. RAW files processed using DxO optics software to normalize the geometry only. No other changes were made. Captioning and contact sheet were done using ACDSee photo editing program.
Hope this helps,
Clemence
Thanks for the effort , clemence!
Youāre right, it doesānt reviel much tint shifts nor the tint shift in maukkas measurements, post #584.
Edit:
What it shows on my screen, though, is a little green content in the sm403, but not much.
Thanks for your advice in post #586. Perhaps 219B sw45k is the way to go for me.
I was also considering using the h17f for a single emitter build. However, as I understand it, these nichia219Cs will burn out at the currents the driver allows. I have not tested this yet, but let me know if you do and it works out.
Correct, the H17F is simply a PWM FET driver, unless you can have it setup to limit the max duty to around ~70% it will drive the LED with over 11A. Which is far too much.
If you limit the max duty it works ok, you will just loose around 10-15% output compared to a true constant current driver.
I think up to 3A it is a linear driver. It uses 8x 7135 for that. Only for half turbo and turbo it uses a FET. See Dr. Jonesā description here:
http://drjones.nerdcamp.net/h17f.html
I think it should be fine when using a non performance cell or a current limited power supply while programming in order to protect the LED.
Correct, it will work in the 7135 mode but the turbo modes will drive it at around 11-12A unless the pwm is limited.
Even with a laptop pull I was getting around 8A IIRC which is way over what most laptop cells are rated for and not safe for that reason. With a GA it still gets over 10A.
The only way I have found to run it on an FET is by limiting the PWM. The TA driver works in the same way and that is what I do with it.
Although I hardly ever use turbo, 95% of the time the 7135 mode is enough.
Is there a good cell to use for programming to prevent blowing anything up?
I had this with a single 219B. It melted my first switch. Turned the LED blue during programming but survived. Iād rather limit things though.
As long as the LED is properly reflowed onto a DTP MCPCB and it has a good heat path they can handle a LOT of power. I have yet to kill a 219C. The output just drops massively and it gets really hot. For example it will put out basically the same luemens at 2.45A as in turbo @ 12A.
If you want to play it safe, use a cheap cell that is not fully charged, say around 3.6V.
Good to know the 219c can handle the higher currents without killing it. I believe I read somewhere that the FET used on these drivers top out around 10A, but that could have been in a specific build with other factors. Iāll go see if I can dig up the source.
I think Iāll give this build a go and program with a weaker cell. It will only be on the FET for a short period of time while programming.
That 10A limit would be the max the FET can handle without melting, not the max it can deliver. Although on a single 219C the max would most likely be less then 10A as it will have more resistance then the normal FETās we use around here.
Ah okay, thanks for the details. Does that mean these drivers shouldnāt be used on quads/triples that can pull a lot more than 10A?
This is what I have in a few of my builds but I rarely use the full turbo mode for extended periods.
That would be a question for the guy selling it (forgot who it was). He will know what the specs on the parts are.
You can be sure that they will pull more then 10A though with a triple or quad 219C.