This is a case where it really depends on the device they’re used in (i.e. the load that it demands from the cell, not necessarily just the voltage). I’ve had them be “useless” at as little as 2.9v with a plain multimeter reading (no load applied) and some that read 2.6v (from very low drain applications). The ones in my digital calipers come out around 2.7v but clearly can’t do their job correctly a bit before that point.
I think most battery testers use a small 50mA load to test batteries but often that will separate good from bad even if they have nearly the same simple voltage reading.
Btw, this is also where the huge difference usually is in the cheap-cheap lithium cells vs. quality big name brand ones…ability to sustain and deliver current. Often the cheapies are just no good at all, but depends on the device.
If you get 3.5A with a full cell then the total circuit resistance must be quite high, about 350mΩ, see my DD calculator .
To get 6A you must bring it down to ~130mΩ, which will be difficult since the mosfet is already arround 60mΩ at 4V drive voltage, and a 40T is something like 12mΩ, plus spring, contact, switch, wire… etc resistance.
You can start by bridging the resistors, but that would only remove 125mΩ.
Yeah, I did make a few tries with some piggybacks, and there wasn’t much gain there. Was thinking of bridging those resistors (shorting) but I don’t think it will do any good.
So I decided to order some of Convoy’s drivers – 6 Amps and other types. These are better designed and known architecture. Will have to make an adapter ring for 20mm to 24mm.
The Amutorchs (SD1, SD2, SD3) are just screaming ‘cheap’: shallow driver cavity (~3mm), thin clear silicone ‘O’ rings, very thin glass lens (1.2mm) w/out AR that leaves artifacts – dropped the light on my linoleum floor and it broke! Then the non-reversing body tube (which isn’t a deal-breaker but would be convenient), and I have my doubts on the switch; they added a standard zinc plated washer to stiffen the movement or for support?
Each replacement is a challenge being not very standard in dimensions and having little room to add in thickness.
Addendum: Checked your impedance calculator, but this escapes me. I have a basic understanding of polynomials, but the values you have in the equation are unknown.
Nevertheless, as you pointed out the internal resistance of the driver and electrical pathway is too high.
And this answers a fundamental question I’ve been groping with – why some of these lights don’t perform as they should.
The values are what Excel gave me when asked to find an equation for the curves (taken from LED tests on BLF), I selected 3 degrees polynomial because that was sufficient for the equation to match the curve, I needed the equations to plot them in the online graphing tool.
Thanks for the added info. Much more sophisticated than at first glance.
But the driver inefficiency is the bottleneck. I have somewhat the same problem with a CULPM1 (HX Boost) with a Convoy 8 A driver. I only get some 1210 lumens whilst I should be around 1800+. I presumed Simon’s driver to be of higher efficiency but it seems not. Or then again, perhaps other factors are at play. I’ll have to do some more testing if the switch or other pathway is hindering the current draw.
It doesn’t seem to be only the driver though, there is clearly some significant resistance elsewhere (likely springs and switch).
Regarding the 8A Convoy driver, it uses a buck converter, it looks to be synchronous (no diode, so no diode drop), if it is capable of 100% duty cycle then the voltage drop (in dropout) is due to the resistance of the inductor, high side FET, Rsense, and also the RPP FET), so like direct drive it is a matter of circuit resistance. If it’s not capable of 100% duty cycle then it has a minimum voltage drop, e.g. most ≥8A TI buck converters have a 86% max duty cycle meaning Vout cannot be more than 86% x Vin, so with a single li-ion cell that limits the max current, yet they can still be very efficient (the drop doesn’t lead to conduction losses).
Regarding the Osram HX and 8A driver, made some tests.
Using a fully charged PLB 26650/5000mAh battery (14.5 mΩ IR - 1kHz) and without tailcap nor springs (button style driver) getting 7.39A and 1370 lumens. Values are on start – not AINSI.
I had previously doubted my lightbox and after much tinkering, this set-up is as close as I can get it – it concurs with most of my throwers and ‘blasters’ (those that have much spill and throw – sheer brute force).
If the driver (buck) is highly efficient, other than the wires, I’m at a loss at this point.
So I’ll have to think of other possible bottleneck(s);
Addendum:
Simon’s CULPM1.TG binning (8RF) should give 1690 - 1900 lumens at 6A.
Chapelle has some terrific skits. His imitation of music artists is his real talent. Prince was to die for. This R.Kelly imitation is fantastic.
Can you believe it… all that money. R.Kelly made a killing… threw it all away. But then, what he did was just so f’ing ugly. He was found guilty on all counts. Spotify ought to take down all his music.
Are you mixing up the binned output with Out The Front (OTF) output? There are always going to be noticeable losses from the reflector and lens, etc…
(Note: I’m not familiar with this led)
I’m comparing the binned values to OTF – ballpark. As the spec’ed values are at 6A and I don’t get near that at 8A (actually 7.39A at tail).
A further test, now with the wires. Removed the MCPCB w/ LED and mounted on static aluminium heatsink (no fan cooling). Driven by a power supply. Current measurement via clamp meter - closest possible to the MCPCB. Test leads 16 ga. On Start (O/S) values.
With the original wires, 24 gauge (I estimate): at 6A –> 1130lm, at 8A –> 1315lm
With new wires 18 gauge: at 6A –> 1230lm, at 8A –> 1350lm
So it would not seem to be the wires that are choking the output. There is a slight increase in luminosity but nowhere the claimed values.
Eliminating the switch, the spring, the battery, the driver, wires, and lens, I can only estimate my LED is either bad binned (which I would doubt as Simon’s are well documented and although a bad one may be in the lot, it is not common) or my lightbox is in error.
The lightbox;
I had problems with the throwers. After some exchanges with Jeff and CNCman, added a dispersing cone and this remedied most of what I had encountered with SST40 and SST70. The Osrams were always reading low, and so remains the problem with these leds.
The box measures well with the Q8 (4 x XPL-2), the Avaritia NA40-SE (3 x SST40s) and the Sofirn SP36 (4 x LH351D). Also works with the smaller EDCs.
Will have to find a better dispersion device for that hotspot…
Hey, anyone watching the youtube lately see an ad for a new movie, where an actor or stuntman decides to “get real” and signs up to be a soldier or something in the MidEast?
The name is something like “Hevic” or “Hevil” or “Helva” or something like that. Keep getting garbage results from goggle, so was wondering if anyone saw it and recalls the title
Searching google does not seem to show too much results. Found it while browsing YTS though. Heval (2021) YIFY - Download Movie TORRENT - YTS. Seems to be a curiosity stream special. “heval documentary” yielded better results.
