Ill add a bit more of my observations if it may be helpful on this project too.
Ill add I have tried three separate types and batches of 7135s. My Candle uses the tint mixing firmware as well so I figured it would be uniquely helpful here. Anyways.
The first 7135s I tried were from Richard and were marked 7135 35L with nothing else on them. The way the tint mixing firmware is by default includes a minimum PWM value of 1. These 7135s worked perfectly fine and had no issues. I can ramp both channels and they both worked and all of the intermediate mixes worked just fine too.
The next batch of 7135s I tried were from LCSC and were the “raptor claw” ones as suggested by Lexel. These particular 7135s had no interested in turning on with a PWM value of less than 5. Set the ramp to a minimum of 5 and they kinda turn on but any mixed value ends up turning them off when we are mixing channels. Even with this, turning on in moon in single channel no mixing then ramping UP turned the light off again then on even though I never set a value of less than 5 in the ramp. Setting the PWM values to around 15 seemed to correct the problem as it would always be on but the ramp was still wonky as in turn on moon would be brighter than a bit up the ramp then it would get brighter again. Made no sense to me as the PWM values were constant. This affected the special modes as well. I ended up throwing these 7135s in the rubbish.
My last batch are Richard’s 380mA with the shark fin. These flicker on PWM value 1 but I raised the the 1s in the ramp to 2 and they seem to be ok. This minor changed seemed not to affect the firmware either.
I’m waiting on prototype hardware before I try to adjust anything about the tint ramping. It could be fine, or it could need some tweaking. We won’t know until after doing some tests.