That explains it!
Don’t you just love that moment in a project when your light gains sentience and starts designing itself?
Mines just developed a permanent standby moon mode on one emitter…all by it’s self. I’m so proud…hah! 
-
Well, I checked to make sure I didn’t have a short between reflector and led -, or somewhere on the driver. So my best guess is that a 7135 chip has failed in a partially on mode. Anyone seen this happen before?
I guess that’s the risk of having the leds “live” all the time and relying on the 7135s to remain fully off when no power is applied to the MCU.
That is also the emitter that had the flickering issues before the LDO fix so maybe that one 7135 responsible for the flicker has been cooked half to death during all that testing. The regulators where certainly getting way too hot during those early test runs so I guess it’s not too surprising.
No way I can indentify or swap it out if that’s the case so I’ll have to live with this on-mode status LED. It’s the center one and everything so it almost looks intentional, haha
Well it’s either that or the light’s jealous of the thing in the background that’s been occupying my attention today 
Oh my god those first 2 sentences made me laugh hard!!!
Damn. Its sb's Borgs at work. I feel your pain.
Did somebody say beamshots?
I’ve paid much more attention lately when running my light and have noticed some flickering also. Thought I’d let you know as earlier I had not. I’m going to try lowering the value of the resistor and see it it makes a difference, at least while I’m waiting on my new PWM-less driver som OSH Park.
That’s very interesting, yes please let us know what affect a lower resistor has. #
And I’m very sorry that all this talk of flicker has caused you to notice a subtle issue that you may not have ever seen without looking for it…very sorry
Haha, don’t be. I’ve learned quite a lot from following your thread. If all I have to do is easily swap a resistor I will have greatly benefited from your issues. I’m going to do the firmware test with constant on/high first, just to see if my flickering can be hammered out without PWM as my PWM-less board is in production. I’ll be adding a lower resistor anyway of coarse, but it’ll be interesting anyway. If it works I could just recode the software to do constant on/high on my highest mode and PWM the others until my new board arrives, but will do both tests anyway. I’ll be adding the test results in my own thread when I get ’em done.
Yes if you’re light is anything like mine, the constant on pin should eliminate/reduce the flicker. BUT at the the expense of lowering the driver performance even more! In my testing the runs with your custom firmware where the worst performers overall. So if MCU current supply is also the issue on yours then simply changing the resistor will solve all the issues in one.
High (100) with fast PWM shows absolutely nothing but a constant solid output when I scope these STAR drivers, confirming what Wight and JonnyC and all those have said. So with the right power supply in place (modified zener or 5v LDO) there’s not going to be any difference at all between setting the pin to ON or PWM 100.
Cheers
Linus
Interesting stuff! I also look forward to your results!
I decided to just try and solve the issue instead of doing the experiment with the firmware (which would result in performance loss), so I soldered on another 200 ohm resistor to get down to 100 ohm (didn’t have anything else lying around). And it solved the flickering… At least for as long as I had it on, which was a lot longer than the time I had to wait to see flickering. I did not re-charge the cells between the tests.
So… my flickering was a quick fix, thanks to you and wight. Basically you guys did all the head-scratching and testing, and I’m just taking a free ride. So I extend my gratitude, thanks guys 
I’ve got 50, 75 and 100s on the way with my next component delivery, I’ll run a few calculations through the zener calculators wight provided and see what I end up on using, but lowering the resistance certainly seems a must.
Awesome! Nice to hear similar results with reducing the zener resistor! So now there’s two solutions to this problem, and your M6 photon blaster just got a healthy boost in sustained output and runtime! 
If you have the time, it’d be interesting to make up a little table showing what zener mod components you need to run a certain number of 7135s. Just theoretically using that calculator and based on the 7135s drawing about twice as much current as is stated in the datasheet. Maybe something to add to the popular stickied zener posts.
Just tying up a few loose ends and completing the handle. Added a 2.5mm jack to feed the clicky signal into the body of the light. Now easy to remove the handle completely.
I actually planned to mount the socket into the tailcap with the male plug on the end of the spiral cable, but I couldn’t really figure out a nice way of doing it without the plug end being a bit messy and the socket a nuisance in the way of removing the main power XLR.
This alternative works quite well and it keeps out of the way nicely in normal operation, but still if the handle needs to come off for whatever reason it’s a doddle.
Other than that I’m sorta seeing the end of this one, the to-do list is shrinking.
I’m still waiting on parts for the blower rig to get here but I’m sure I can find something that still needs doing in the meantime…like some beams that need shooting or something
Cheers
great work!
Where’s the beam shots??
Still looking plenty stock & sharp.
Where did you get the coiled cable if I may ask?
You want a beam shot here’s a beam shot.
Happy? :bigsmile:
They’re coming though… Just taking a while cause the light is still being worked on and it’s often not in a workable state. For example at the moment the battery pack is in pieces again as I’m swapping over the relays for mosfets with some control circuitry.
Thanks guys!
I assume you mean the small switch cable and not the power cord? For the small one look on Ebay, just search spiral usb cable or spiral/coiled charger cable or something like that. Lots out there of various lengths and connector types.
Sorry, I forgot that your design is with an e-switch, the coiled cable is not carrying the juice.
Technically it’s not an e-switch. It’s a reverse clicky connected to the MCU side of the driver circuitry with the coiled cable.
But you’re right it’s essentially like an e-switch and the small coiled cable only needs to carry the 30ma or so that the MCU requires to run.