Made a bet with them and bought some. Since theoretically they can deliver up to 8A to the emitter without brakes, I modified one ∅17mm unit by swapping its R005 sense resistor with a stack of 2x R020s. An unmodified ∅22mm driver I have installed in a C8S does just a smidgen above 7.6A, while the modded unit measures ≈3.875A max (DC power supply amperimeter readings). So it works, but something is not there. The unmodified driver blinks 36 times when ramping, the modded one also does 36 but only the later 30 can be seen, there's a brief pause in the darkness when it starts from minimum. Also, the special “3rd” mode selection blinks remain }P unseen and the 1% mode doesn't works (unseen too).
I do not have a good explanation as to why this happens, it shouldn't happen as V = I × R and hence the sense voltage remains :???: constant. The special mode selection blinks are faint, they may be using the 5/36 or 6/36 ramping brightness level and this tells me if I stack more sense resistors they will eventually become viewable. But what about the 1% brightness level? :facepalm:
That other SST-40 5A max driver with sequential mode change is just another story. Did someone try sense resistor adjustments with that one? The lowest mode could also stop working with it, even if just a little reduction shouldn't be a problem.
Anyway, what matters here is why in @#$% does this happen with this 8A weird firmware units. Sure there is a reason somewhere. I have another ∅17mm weird firmware unit waiting to be tested with less than 10mΩ of sense resistor, so we'll see.
The second ∅17mm 2x selectable modes current with ramping plus special mode driver I modified worked as expected. I swapped its R005 with a R010 + R050 + R200 stack, and it ended up delivering 4.75+A; used it to power a Convoy M1 with Osram 1mm² White Flat.
A 90.2? SBT90 gen2? Setting sense for 12.5A means very hefty increase in power dissipation at the MOSFET (roughly +50%), good luck with that. Try to increase MOSFET cooling somehow, thermal glue can help. Stars-922 here and there, Tian Mu grey (this one starts pretty liquid) or just find or use the one which you like best.
Does anyone think it’s possible to mod these for 2s input into an XHP70.2, like the zener mod on old 7135 drivers? That may be a silly question, I don’t know nearly as much about circuits as you. Thanks for experimenting with these and posting results.
mountainair26, me and others had a related conversation concerning the components in the newer Convoy drivers in the Convoy XHP35 Driver Analysis / Testing / Schematic thread. Long story short here, after I found that their onboard linear regulators work up to 18V of input.
Bear in mind the power dissipation at the onboard MOSFETs still matters. Since they're temperature controlled, they should work straight away with 2S cells driving a 2S emitter. However, they are likely to step down current aggressively, due in part to the larger input to output voltage difference. Reinforcing MOSFET/driver cooling amap and removing the onboard temperature protection is going to be key for proper 2S emitter driving, imho.
Also, unless something has changed the ∅17mm SST-40 5A 12 groups driver has a problem, with missing upper outer track (solder mask covered) which is key for cooling. Check here and my solution there among that maze of nested quotes.
So, a (XHP50n/XHP70n) 2S emitter and 2S battery S2+ build is feasible without much difficulty, for example. The only problem could be removing the heat from the driver quick enough. A copper pill can help, of course. Simon made a run of copper pills for standard S2+ flashlights/hosts, he told me to order using the current copper pill ad for the square threaded copper version, and send him a private message requesting copper pills for alu S2+ (check here).
Double thanks Barkuti, really appreciate it.
Does this driver from the nightwatch NL40 seen in Kawiboy’s post #322 look like it’s similar enough to the convoy to mod for current and/or remove the temp - throttle as people have done in the convoy drivers? What do you guess?
The driver looks to use the same technology as the Convoy drivers, maybe it comes from the same OEM too. If so, the NTC resistor to remove is at the right side of the MCU in the above photo, below that regular resistor. The photo lacks resolution to say anything else, but that regular resistor above is probably in series with the NTC, and then connected to pin 8 on the MCU. In the Convoy drivers the NTC is connected to pin 8 at one side, and to ground by the other side. Using a resistor between the NTC and the MCU's pin 8 means a form of calibration is obtained, dampening the sensitivity with increasing value. I also thought about this while tuning Convoy drivers, but they don't have dedicated onboard space for this.
Been looking at THIS driver for a single SST20 in a Convoy S2 w/SMO, think the LED will be alright with it?
(Currently have a QLite driver in it, but I’m not happy with its limited modes and power output, adore the combo of throw and flood tho!)