Thanks TK. Yes this particular example does have the voltage divider removed.
I also have another 17DDm like driver where I plan on swapping to a tiny85 and converting to eswitch, ill be sure to remove the voltage divider resistors.
Sofirn C8F where I want to drill some holes and put a ring of Yuji in the blank space in the reflectors. This is a bit of a long term project especially since I need to mod the driver a bit to enable independent dual channels (cut trace), and modify the host as well. That and work
You could also maybe drive those from the aux LED pin, using the existing facilities for that. Maximum power is lower though, like maybe 30 mA instead of 350 mA. And it’s not really dimmable except for high/low/off. But, for designs like this, I recently added some support for making the aux LEDs and main LEDs mutually exclusive, so it won’t have both on at the same time. It would still be a little awkward though, since that would mean the aux LEDs are only on while the driver is asleep, like in “off” mode or lockout mode.
Ideally I want to drive the Yuji off the 7135 channel, changing the ramp to accommodate them. They will be a pre throw warm flood. They will also serve candle mode.
The niggle with them is they really only like 30ma each so the PWM values will all be low. I am guessing MAX_1x7135 should set the max 7135 PWM value for the strobes?
The MAX_1x7135 value determines which ramp level is the highest before the FET activates. It does not determine the PWM value.
If the emitters can only handle 30mA each, and there are three of them, it may not be a good idea to drive them with a 350mA chip. Even with PWM to reduce the average current, the momentary current will still be about 4X the safe level.
But if you try it, it looks like the 7135 part of the ramp should max out at 65/255. (3 * 30mA / 350mA) * 255 = 65. So, after calculating a ramp, multiply all the 7135 ramp values by about 0.257 to scale them down to the desired ceiling level.
Or you could maybe replace the 7135 chip with something lower-powered… or put a lower-powered chip on the aux LED pin and treat it as a 3-channel driver.
Do these drivers have the voltage divider resistor? There are two resistors clearly shown on the fourth picture. It it near the 7135 chip or on the opposite side?
Finite State Machine, or [F-word] Spaghetti Monster.
Why? Because it takes the mess of spaghetti code typically associated with low-level firmware and locks it away in a box where people don’t have to care about the monster inside. Instead it presents a more user-friendly interface in the form of a finite state machine, because that’s typically how people describe flashlight UIs. A finite state machine is basically a flowchart.
Yes, the one labelled R1, connected to pin 7, is along the path which allows current to leak. Removing it would stop the leak.
For BLF-A6 and Bistro, this is a good thing. The offtime measurement depends on current leaking out at a predictable and high enough rate. And for multi-cell serial lights it’s necessary to scale voltage down to a usable range. But for single cell e-switch lights it’s not desirable.
Those banggood drivers don’t have a tiny85 though, so they can’t run FSM.
Thanks TK for all the explanations. I have put Attiny 85’s on a couple of these drivers and run the eswitch between pin 2 and earth.
When I read through these posts I feel like a Friggin Scared Moose.
I figured that what else it could be while laying in bed thinking about this. Too far into this hobby where this ends up happening…
I would definitely need to test this on the bench with the PWM to see how it behaves. I have seen a test of these where at 120mA they start to go blue… Are there lower power 7135s?
Lots of code updates lately, though most of them aren’t UI changes. Most are to keep code cleaner and more manageable, splitting hardware-specific bits out into their own files and such. This way, new drivers can be added more easily with less clutter. Drop in a hardware layout definition and a UI config file, and it should be pretty much done.
Due to the way the C preprocessor works though, it does typically still require adding a couple lines to the main source files. But the changes there are a lot smaller than they used to be, and easier to read.
Anyway, I paid off some technical debt so I’m done for the evening.
I wonder how it even compiled like that. I’ve been using a script lately to build every supported version every time instead of doing just one, and the Q8 version built and worked normally under gcc 4.9.2. But it looks like it really shouldn’t have worked…
Since it’s a thing people might want, and since I think Lexel may have been requesting it, I took a moment to make a Werner-style momentary UI, side e-switch plus tail clicky-switch.