Wellp, I was going to submit a P4000 review in Kaidomain's site, but it complained “Warning: Review Text must be between 25 and 1000 characters!”. It also complained about name too long (max 25 characters), LMAO. My review is well above 3K characters long LoL, so here it goes:
With a sense resistor stack consisting of 50mΩ plus 1Ω sense resistors in parallel (47.619mΩ combined) and driving currents of 4 and 5A in high and turbo modes, by Ohm's Law (V = I × R) and electric power (P = V × I = I² × R) formulas it is easy to infer the voltage and power drops across the sense resistor stack: 190.476mV / 761.9mW at 4A and 238.095mV and 1190.5mW at 5A. These are HUGE figures, causing a lot of wasted heat and a massive drop in efficiency.
When a typical AC led driver which outputs many tens of volts uses a ≈0.2V sense voltage, the impact in overall efficiency is no more than very small. When a driver meant to output a little above 3V uses ≈0.2V sense voltage, the relative impact is huge with ≈6% output power loss (!) just in the sense resistor. This is gross and inadequate in a driver like this. Lazy job from the driver designers.
Still, a driver which does its job. For 2 cells in series can be adequate.
Here it ends the text I submitted in the Kaidomain's page (limited to 1000 characters). I also add the following:
Edit.- Of course, by Ohm's Law the sense voltage gets lower in lower modes (V = I × R), and so the negative impact of it is proportionally lower in these modes. But still, lazy job tu use such a coarse sense resistor stage. Good drivers employ just ≈50mV (≈0.05V) maximum (which can still be improved by using multiple sense resistors, namely as the current output raises).
I do not mean to say that the driver cannot work, but the absurdly high sense voltage (even if not much in practice) unnecessarily restricts regulation time when powered by single cell voltage, and the increased heat dissipation on top of a (undesired by power users) “smart thermal management function” means the stuff is ultra-likely to taper down the current output rather quickly in high modes.
Single cell overdischarge cut-off, at 3V, is high. Considering that in a flashlight the driver does not see actual cell voltage, but after it goes through springs and switch losses, actual cell cut-off happens above 3V at the cell terminals; also, considering that most li-ion cell manufacturers specify 2.5V as cut-off voltage in their cell datasheets (and down to 2V in a few exceptional cases), there's no good reason for a 3V cut-off. And yes, I know that what happens at 3V is the stuff starts flashing the led (low voltage warning), with actual, real cut-off happening at 2.7V. But this is senseless, stupid if you ask me. Just imagine for a moment that I am using the flashlight for a walk or jogging: do you think I'd be happy if the flashlight starts flashing the led while I am running at night? Certainly NOT. I'd very much prefer the flashlight low power blinking just a few times to signal the low voltage warning, and then dropping to the lowest available mode. In the lowest available mode, the flashlight can still serve me while it discharges the battery down to let's say 2.5V or even 2V (so, effective cut-off at 2.5V or 2V). That can be a useful lifesaver while I am walking or running at night; blinkies… not so much.
So, in my opinion the low voltage warning can be set at 2.8V or 3V, with a few blinks to signal the warning, then dropping to the lowest mode and continuing the discharge down to 2.5V or 2V. At the lowest mode (just 5mA output in this driver), the driver could still provide a good deal of light time while it hits 2.5V or 2V. In any case, a standard white led (blue pumped phosphor converted led) has a minimum Vf of 2.5V, and this means that while powering a standard white led the led poses a hard cut-off limit: 2.5V. Red or other types of leds have a lower Vf, but this is another story.
All in all, my personal evaluation for this driver is 3/5. Which isn't bad, but considering how fearful these fellows are of anything below 5/5 or even 4/5, we'll see if this (the above part I mean) gets published in Kaidomain's site. 
Thu, 03/04/2021 - 14:56
