The fet provides as much current as the battery can provide for turbo and the +7 provides regulated output for the lower modes usually via an Amc7135 current regulator
The 8 AMCs are split into 2 banks, one bank with a lone chip and the other with seven chips.
At low levels it drives the LED through that single AMC.
At medium it uses all 8 AMCs.
At high it switches to MOSFET.
Overall that’s a fairly simple, cheap and compact design that delivers a lot of power. It is a bit more efficient than the more common FET+1 drivers but not as efficient as the more sophisticated buck and buck-boost drivers.
Basically direct drive, or maximum power, available
Relatively simple driver style
Very easy to make single-sided drivers, even in small form factors
Disadvantages:
No regulated output
Relies heavily on PWM for all non-100% modes
Brightness goes down as battery goes down
Common single-channel AMC7135 linear driver
Advantages:
Simple driver
Regulated, constant-current output (until battery voltage gets lower than Vf, that is)
Disadvantages:
Low modes still rely on PWM
When battery voltage drops below Vf, brightness decreases with battery voltage ("drops out of regulation")
With high AMC counts, this is more difficult to get single-sided drivers with
Two-Channel Drivers
Advantages:
Two styles: FET+1 and 1+7 (not always 7, but that's the standard)
the +1 gives a regulated output mode, and also allows for a low moonlight mode with the minimum brightness multiplied by one AMC7135, compared to eight for the more common single-channel drivers
If FET+1, still also allows max brightness, and takes many of the pros of FET driver
If 1+7, allows a standard high modeof around 3 Amps while preserving the lower moonlight of the 1 AMC7135 (see: Moonlight Special)
For FET+1, often easy to make single-sided drivers still
Disadvantages:
More complex driver design
1+7 still hard to fit single-sided (as with 8AMC7135 as well)
Triple-Channel Drivers
Advantages
You get lots of output options.
Full power with the FET
Regulated High mode with the AMC7135s (both 1+7 channels or just the 7 channel, depending on configuration)
Low moonlight with the single 1xAMC7135
Higher efficiency than FET+1
Disadvantages:
Very complex driver design and complex programming, too
Seems to barely fit at all on smaller form factors like 17mm, definitely double-sided
Note: For basically all drivers besides the straight-FET drivers, you can add or remove AMC7135 chips to change the current. For example, in one of my smaller lights, I have a 7+1 driver with two of the 7135s removed because I don't see the need for higher power that the battery and host can't really deal with. I've also got a triple-channel driver in ~46mm for a different host, and that's 1+15+FET, because the batteries, host, and emitters can all handle that under regulation (it's also single-sided because it's so large!). I've seen single-sided FET drivers in 15mm size, while I can't think of a single-sided 8x7135 driver in 17mm or even up a couple sizes.
Hopefully this has been informative. We were all new once ;)
I just flashed Narsil on a Texas Avenger driver, but I don’t have the perfect flashlight yet. only on a test station. It was my first time flashing an eswitch firmware and I believe it worked out. I believe Narsil has 150 steps for ramping.
What if you have an xpl hi rated @ 3amps, a mosfet+7+1 and a 30Q battery, will this mean that all 15amps of the cell be unrestricted due to MOSFET and fry/damage the xpl hi?
No, the XP-L HI will draw between 5 and 7A and won’t fry (if reflowed properly on a DTP board of course).
Current is not pushed by the source, it’s pulled by the receiver (LED)
EDIT : I should add that current draw is also linked with the Vf of the LED… A low Vf LED (like the Osram White Flats) will pull more Amps than it can handle from a high drain 18650.
Thus it’s not recommended to go full Direct Drive with these kind of LED.
The XM-L2, XP-L HI/HD, XP-G2/3 are safe in DD though
Nice way of explaining that. But not fully correct because it may make someone think that a 1A LED will pull only 1A on direct drive.
So maybe it would be better to say:
Currect is negotiated between the cell and the LED.
Ah interesting I see. This means that the 5-7A is the max the emitter can pull unrestricted. I had thought that the amps the manufacturer suggested operating amperage at 3A is what it pulls but that’s just a recommended suggestion to limit the “throughout” of amperage using MCEs 7135 or via software limitation by firmware?
So does this mean that my qlite driver for the xpl hi v2 3a rated at 3A can add a few more chips to allow 5A throughput to increase emitter output?
Yes this is a common way of increasing current in a regulated 7135-based driver.
Stacking them is a work of art, and it allows you to control the max current.
Oh, and we don’t follow manufacturer’s recommandation down here… We tend to push the LED TO THE LIMITS :smiling_imp:
Damn, so I severely nerfed my first c8 with the qlite driver then, its operating so conservatively right now that I feel disgusted I’ll give it to a friend as it’s likely going to be stable for life and user friendly.
On the other hand, I did order 7135 chips, perhaps I can try stacking.
So, if emitters pull the amperage, it’s up to the modder to determine how much an emitter is allowed to negotiate and this is what people tinker with to overdrive the emitter to its limits?
the current through the LED is determined by both the LED and the battery.
if you have a really good battery, that drops very little voltage when providing high current, you can fry your LED, xpl, xm-l2, xp-l2, nothing is safe with DD driver.
That’s not true. Even the most powerful cell won’t deliver more voltage than what it was charged to. Which typically means up to 4.25V.
And even with 0 voltage drop in the cell and in the host - 4.25V is not enough to fry some LEDs like XP-L or (I think) XM-L2.
Taking into account resistance which is inherent in cells and hosts - there are many more perfectly safe DD setups.
Though nowadays this DD safety is becoming scarce, most recent LEDs can’t be simply directly driven from a powerful cell. Though they still can from a less powerful (f.e. smaller) one.
sorry i was using bad examples, but my point is still valid, the current through led is deteremined by both the led itself and the battery.
in my modded emisar d1s with osram white flat led and stock driver( DD driver), using high discharge battery will fry the led, but using something like samsung 26f is fine
Something like an XM-L2 or an XP-L HI, or an SST-20 70CRI/LH351D 70CRI will have a high enough VF to prevent damage to the LED, even from a perfect 4,2V source.
The problem comes from lower VF LEDs, like the 95CRI SST-20s, or SST-40, Nichia 219C 90CRI.
With powerful enough cells(15A+) and a very low resistance circuit(BeCu springs, dual copper alloy springs of any kind, spring bypasses, very low resistance path and components), an LED can die.
This is why something like an FT03, with its dual phosphor bronze springs, its SST-40 to the absolute max.
With a spring bypass/dual beryllium copper springs, there is a chance you could kill it.