Would you mind sharing the code (did you modify the qlite-code itself)?
I found the qlite version very good because of its 3 V cut-off, no other ’7135 seems to have that.
Greetings
The pre-rev.A Qlite was 20mA - 2% - 25% - 100% on the 4th group.
The driver/code that I was using for the moonlight mode tests was for my custom SkyRay King driver… not compatible with a Nanjg/Qlite driver. Custom SkyRay King SRK Driver I am assuming that the troubles with the Qlite in low modes is due to the same issue.
Holy Toledo, now that’s what I call a driver…
But thank you for your answer, I’m experiencing the loss of ‘moonlight’ as well, although it’s not permanent, yet.
When Hank of IOS had first leaked information about a revision of the qlite driver after the nanjg-incident there was some talk about a 10mA - 2% - 25% - 100% version. I was a bit disappointed when it turned out 5mA in Rev.A, I found 20mA too bright and 5mA too low - a matter of taste, of course. But loosing the very-low completely isn’t really helpful…
I still like the qlites for many reasons. Perhaps Rev.B will get us a 10mA - 2% - 25% - 100% version 
Greetings
Do I understand that correctly, the moon in NLITE doesn't light up the LED?
I used a quite conservative level of 9/256 at 18kHz; I never had problems with that value.
If needed, I can make another NLITE with a slightly higher value.
HarleyQuin: What bothers you with NLITE?
I am using 10/256 as the lowest setting in my custom SRK driver (Custom SkyRay King SRK Driver) Works fine with all the LEDs that I have tested (even after some considerable heat soaking tests) It puts out considerably more light than a moonlight mode… plus I am driving 24 ’7135 chips. I don’t see 9/256 as causing any problems for a eight chip driver.
I started out with values around 4/256, but as the driver was put through its paces the required value needed to reliably light the LEDs kept creeping up as it aged and burned in. The threshold shift was permanent.
It's only happened I think 3 out of +20 builds, and always with an LED that had been abused in some way. I don't think you should change anything just to cater to a few damaged emitters. :)
I ordered 2×Qlite drivers from IO (2 month ago) for my personal lights.
I installed one in Jacob. I changed wires for 22AWG ones, changed contact spring for bigger one.
Default modes were working good except “Memory function (sets in after 2 seconds)” - that is not true is more like 5 seconds.
It says “No visible flicker on any mode”. Ok it really worked without visible flickering for a week(20 minutes per day). What happened?
Today I turned on the light and it has high mode only. Flickering high mode.
Well I tried different switch and same thing happens.
I am not expert so I am wondering what could happen?
Anyway I was so upset that I decided to swap it with another one. But this time I have left stock wires and spring.
And this time everything worked perfect for 30 minutes so I thought ok maybe I messed something with first one and minute after that thought again same thin happens it went for “high flickering mode”.
I can really not explain this. Perhaps some witch cursed me?
I don’t have nothing against IO and I spend about 200$ so far in their store. They have good stuff for business but I only know that I don’t recommend this driver and I will certainly not order this p o j again.
I installed more than 15 single mode no name 8×7135 drivers to my friends/costumers and not a single problem so far. I even decided to leave stock 22AWG wires on them.
http://www.aliexpress.com/item/17mm-7135x8-MCU-Dimming-2800mA-1-Mode-LED-Circuit-Board-LED-Driver-for-CREE-MCE-P7/724554906.html
No flickering with this babies and single mode only.
By 'no flickering' it means very high PWM, not 'no flickering' as in it will compensate for a bad/loose connection or short. Being stuck on high mode is the classic symptom of the LED- shorting to ground. When it does that, it's running in direct drive and bypassing the driver - if it's an intermittent short to ground at the LED-, it will flicker when it switches between regulated 3 amps, and direct drive.
Thanks comfychair,
XP-G2 de-dome, noctigon star(stock I did not re-flow it)
Tailcap readings are 3A on fresh pana ncr18650pd.
If this is really case of led- shorting to ground how come that this happen after week of use or like in last test after 30 minutes? Strange.
I changed modes constantly during last 30 minutes test and they were working fine and suddenly high flickering mode appears
I don’t know how I managed to short the LED-?(I did this like any other one? Artic alumina)
Anywhere in the driver LED- circuit shorted to ground will cause it. That could be a too-big solder blob shorting to the reflector, a wire that's too tight up against a sharp corner and when it gets hot the insulation softens, the point on the driver where the black wire attaches could have too much solder and short to the inside bore of the pill...
Run it with the driver removed from the pill and no reflector, and reassemble ONE PIECE AT A TIME and see which part causes it to start acting weird again.
Thanks on very useful advice. I will have to do all that. I am to disappointing today.12PM here.
I will have to do that with fresh head and nerves
I have custom made testing XM-L2 LED with aligator clamps but I did not use it in this setup because I was to sure in myself.
I will report tomorrow after I de-solder Qlite driver.
I have a Qlite driving my de-domed XP-G2 in my 612 from FastTech. Everything worked great for weeks until one night I went for a walk just to see how the throw was. I never ran the light past a few short bursts, maybe a minute max. During the walk moonlight mode started flickering then stopped completely and the next mode (low) started flickering. The light works flawlessly on medium and high.
...and I have used probably two dozen of these things, and each and every one I put through a torture test which is to drop in a freshly charged cell, switch it on high, and tailstand it on a table until it shuts itself off when the low voltage protection in the driver kicks in. I have not had a single one do any of the things you guys are describing.
I have another one that is on my high CRI XM-L2 in another 612 that I have. It use to go in moon light for about 10 seconds then got stuck and would not turn on. Then I desoldered the star and it works perfect. I decided not to try moonlight again on that one so its just low, med, high now and works great.
my xintd c8 v4 doesnt have any problems using moon mode after being in high for a long time
Now, mine is also flickering on high as well as low. It’s one of my XML2 that shuts down early at low voltage DD (flickers just before it dies). It’s not an even, PWM type of flicker but more erratic like a dirty contact. The switch is a hair trigger when changing modes so maybe that’s it. It’s barely noticeable so I’ll live with it for now and see what happens.
Texaspyro wrote
What, millions weren’t spent developing this chip just so that we could come up with new ways to break it?
millions of zimbabwe dollars perhaps
You guys are not alone.
I did a T20 mod last week and two Qlite drivers have been swapped for this light and I had the exact same problem as yours. Works well initially, and then after certain period of time the light starts to flicker in all modes.
At first I thought it was my soldering problem, so I swapped it into second Qlite driver, and the problem persisted after two days of normal functioning. So far I have 4 Qlite drivers with flickering problem in my T20, T08 and C8 mod. I even created a thread asking for help on my flickering issue, and comfychair adviced me to check if anything bypass the driver. I did, I reassemble everything carefully and it works like normal for two days before it happen again. I understand comfy’s point but I just feel weird with the driver. Last night I swapped my T20 driver again into its stock driver and it works fine now. But I still need few more days for observation.
Coincidentally all of my mods have stacking of 7135, but I did solder them well.
Not trying to blame Qlite directly but I starting to feel strange with it, not having proof nor test though. I will try to order some ‘normal’ Nanjg 105C from FT in the future for more experiments.