What did you mod today?

“Made” some programming keys :blush: and got Anduril working on a FT03.

Added a FET driver to my Convoy L6 with XHP70.2 in 5000k, the 30mm 2S Momentary driver from MTN.
I love the spring bypass all the way through to the MCPCB :smiling_imp:

Dam this thing gets hot, almost heats up like a FW3A…haha
Im running 2 x unprotected 5500mah Keeppower rated up to 15a.
This thing is a serious fire hazard. Doesnt look like ill be letting anyone else touch it and only pull it out to wow people i think.

I turned it on for the first time outside (just in case) and i could see the beam in the midday Australian direct sun, now thats some light.
Cant wait for the sun to go down now :frowning:

I like the SW30 9050 a lot.
It looks better to me than the 9080 in these single samples
they are very close, and things that affect the difference includes, how bright the light is, the type of reflector, and variations between LEDs.

I have not seen anything to suggest to me, in actual use, that the sw30 9080 is “better”. I think the variation between LEDs is a more dominant factor, than the CRI R9 of 50 vs 80

I was surprised. I expected the 9080 to kick ass and take no prisoners, but honestly, I would be just as happy, or happier, with the 9050, in this particular comparison.

the control shot surprised me (LEDs are not even close to sunlight, even at 9080):

Allot of things can be at play regarding the color appearance of daylight in a room not the least of which thought of is the glass. From Arch Lighting.

“Color By definition, the color of daylight is essentially perfect, with a color rendering index (CRI) of 100. However, almost every glazing system has a color characteristic, from very subtle hints of tint to strong variations of green, gray, bronze, aqua, and blue. Remember, just like any filter, the transmitted color is the opposite of the color that is most absorbed. Blue glass, for example, absorbs yellow light. While the color impact is often slight, a heavily tinted glass will cause the perceived color of daylight to change considerably, reducing the CRI of daylight considerably. For example, some green-glass systems have a CRI of 60, about as poor as old-fashioned “cool white” fluorescent.”

Also, higher efficiency glass cuts off a portion of the red end of the spectrum near the infra-red. It helps to explain why, in the Mandalay Bay hotel, why everything in the rooms has a strong green cast, it’s in the glazing. Other factors are also at play like that affect the color temperature of daylight like the time of day the picture was taken, prevailing weather conditions, and latitude. Then, what’s the white balance?

yes, lots of variables
that is why I post multiple photos, to compare to each other,
in the same setting of variables

Im not even saying that the photos look exactly like what my brain says Im seeing. That is a fools errand. So I just post photos to compare to each other. I think that helps eliminate some of the variables and gives a reasonable basis for Comparing one sw30 9050, to one sw30 9080.

Did you measure the LEDs to verify they are actually 9050 and 9080?

Im sure they are what I said they are

sent you a pm w more details, so as not to hijack this thread,
any more than I already have… lol

I put a royal blue G3 in a C8 just for funsies.

I take it in the top photo above, you meant “9080” for the label on the right. But they do look very close. The other thing about a control shot… is the angle of the sun. High noon versus horizon will yield different tints (neutral white vs. warm). But you’re right that there is also the matter of context subjectivity (the color of the object, its light reflectivity/absorption, and environment).

It’ll be interesting to see what emitter bins come down the road in the future. Imagine having 219B 9080 sw30 output tint at 50% greater efficiency!

Jon, are your images taken with AWB? I guess at least the beamshots where taken with a fixed WB at around 5000K, right?

Nice! One step closer to a lightsaber.

Jet-II Ti Anni-Edition. Just arrived, has several flaws, but still … I hate CW.

Very little contact surface

Heating. The solder melted around 190 °C. Means that’s leaded solder?

Nichia 219B sw45K D220

Looks ok to me

The beam pattern … well …

Nice color :slight_smile:

I hope nobody minds if I ask here but is there anyone who would be interested in doing a mod on commission? I have an Acebeam EC35 II, Nichia 219C edition, I was tinkering with the reflector/centering ring trying to fix the donut hole and somehow the LED popped off. Not sure if it wasn’t soldered all that well, the LED diode came off the pad so I think it’s toast. I do have a spare XP-L HI 3D LED to swap in. My hands are just not steady enough to work on small electronics anymore. Acebeam said they’d fix it but I have to ship back to China; reliable shipping would cost more than the light is worth.

Again hope nobody minds me asking on this thread, thought I might make my own post but didn’t want to add clutter. Feel free to PM or reply with your price if you’re interested.

PM incoming

Installed sw45k into Hanko D10 Ti for a friend. What a stunning piece of machine work. The threads are like butter, the finish work is amazing.

Someday ill need to take a stab at reflowing. Im just not comfortable with it.

I’m lurking here a lot, but I’ve never put my mod here before as I don’t think they are worth mentioning.
However I’ve updated my Emisar D4S with FET+8+1 recently and found some weird/interesting behavior of AMC7135s.

Driver is running Anduril. This is crucial information for the story as some AMCs have problem with ramping on low PWM levels. I know that 7135 are different and the ones with claw logo shall be the best for flashlights. Well it depends…

I have bought few AMCs from different aliexpress shops in search for “claws”. Finally found some with number 1937. This is different from the ones that were used on original D4S driver. Emisar driver is using ones with number 1744. I have also “boats” with symbol 35R.
Assuming that boats are worse than claws I have put them on rev 1.1 of my driver (green from JLCPCB) and flashed it. Only issue with it was that boat need higher PWM level to start working and on floor ramp set to 1 it did not lighten LEDs at all. It was easy to fix by setting floor to 3 in Anduril.
Claws were saved for the rev 1.2 driver that I’ve ordered from OSHPark. The main difference between v1.1 and v1.2 is 1x7135 placement. I’ve moved it to have it on front side if I decide to have single-sided driver with FET+3+1. Logically it is still the same driver. After putting driver together and loading same software I noticed weird behavior. Those AMCs did not require same PWM level to work as boats, but I think that problem with 1937 is much bigger then with boats. Basically they do not behave consistent to PWM levels. I mean on lower PWM they are providing enough current to lighten LEDs, but when ramping we are expecting more and more current to run trough them. Nope, 1937 are not working this way. There is a dip in output current somewhere between ramp level 10 and 20. This is really disturbing as during ramping I could see power raising and suddenly LEDs were on lower level again for the moment. Same was observed when second N-channel was attached. To mitigate the problem ramp configuration would need to be changed to start at about level 20 so it is even worse than with boats.
What I’ve actually did to solve problem was replace N-channel with boats and use AMC 1744 from D4S driver for 1x7135. With this setup I’m not able to see any jumps during ramping.

I think 1937 can be seen in this chart from one of aliexpress user:

Here are some pictures of my drivers:

I felt the same. The turning point for me was this hotplate (pic is a link)

Mine’s still working like a champ!

So much nicer than my frying pan. I touch the rim nearly every time :person_facepalming: