Now that Int’l outdoor has the MT-G2 on a PCB, I am thinking of trying to direct drive the MT-G2 directly from 2xLi-ion batteries. Basically, I’m thinking to just mount the MT-G2+PCB onto a pill, and use a contact board and wire directly from the contact board to the +/- on the emitter PCB (noctigon?).
I’ve seen that this (direct driving) has been suggested at a couple of places, with, possibly a resistor for current limiting, but I was wondering:
- What kind of value would be suggested for the resistor (ohms and wattage)?
The resistor would need to be in-line with either the + or - lead from the contact board to the emitter, so, physically, how to do that? Do you just solder a (possibly large) resistor in-line with one of the leads from the contact board to the emitter?
This is more of an experiment, and I’m mainly looking to have the MT-G2 light up and not get killed, i.e., I’m not looking for constant current over runtime.
If anyone has info re. the above questions, please post?
I would suggest getting a host and driver rather than trying to fudge something together tbh, its a very expensive experiment and at the moment, no one has a clue what they’ll take.
There is a host on cnqg that I think, but dont know, will take the mt-g2, the pill seems big enough to take the interesting drivers and it’s 2 x 26650 so there’s plenty of capacity available.
Understood, but assume for the moment that I am willing :)…
Then, say the max voltage was 8.4 volts, and the MT-G2 Vf is 6 volts, so we want to drop ~2.4 volts.
If the current is 3 amps, then v/i => .8 ohms?
So, if I put a 1 ohm resistor (of sufficient wattage) that’d bring the voltage down to somewhere around the Vf, maybe slightly less, correct?
3 amps x 2.4 volt => 7.2 watts, so I’d need ~ 1 ohm, 7.2 watt resistor, that’s small enough.
Does that sound about right?
Just trying to get a “sanity check” (ok, left myself open :(!) for this…
Jim
BTW, I really think that some of the estimates are kind of on the conservative side, e.g., 2xLi-ion batteries might be ~8.4 volts O.C., but it wouldn’t be 8.4 volts, under load.
well that’s the problem - once you get voltage sag, your calculation is out. once the batteries run down, your resistor is wrong etc. It might be fine, probably will be as I dont get how the driver drops the voltage, tbh, I suspect it purely limits the current but that maybe entirely wrong.
My main concern at this point would be to not kill the MT-G2, so if it’s under-driven or sags after ‘x’ secs or minutes, I’m ok with that. That was why I mentioned about the calcs being conservative.
I can always swap in a bigger or smaller resistor (or use a pot), but I would rather not lose the MT-G2, if it’s possible… It’s not been a lot, but I’ve killed a few emitters (and lights) already, so I’m kind of past the first “oh C…P, I killed it” phase :)…
well, just use thin wire to start with and no resistance improving mods and see what happens, I’m on a different continental plate so I feel pretty safe tbh J)
Edit, also remember what your reading at the tailcap is not what the emitter is seeing, it may be low, maybe high, not sure that one 6v emitter has the same double the cell half the tail current relationship of say an xm-l
Direct drive off three cr123s in series is reasonably bright. Or use low-capacity cells like 16340s where there will be enough voltage sag to not kill anything. For any respectable drive currents or usable runtimes you gotta have a driver, that's why they exist.
Am I mis-remembering, or did you post about doing something like this? I searched, but probably either used wrong terms, or missed it, but I thought that you had?
I ran one from a 7135-based driver with a hack to cut the voltage in half for the MCU, it was fine with 16340s (lots of sag meant less voltage for the 7135s to burn off) but was not fine with strong 18650s (not enough sag) and the driver ran very very hot. It didn't fry anything as I shut it off before that, but as it got hot the output current started falling off a very steep cliff and was clear it wasn't going to work without some changes.
If you want to use this in a light, save yourself the trouble and get a buck (sometimes called 'step down') driver that will run from three cells in series. Lots of choices available and they don't have to be listed as 'MTG2 driver' to work. To drive it fully and have the regulation work the way it should so you have steady output through the battery's full range you have to have 3 cells, as with only two cells the input voltage will fall below the point where the regulation can do its job very quickly, as in cell voltage falling to only 3.8v causes the regulation to stop working, which is nowhere near acceptable usable range. The only possible catch is, good buck drivers that do 3-3.5A and up are big and none I know of in the 17-18mm range, only resort to the cr123/16340 for bench testing (and then, you can get enough light to see if it works or not with only two cr123 which is perfectly safe and only runs at about 450mA edit:650mA, sorry!), or if it just absolutely has to be in a p60-style light and you don't care about it being practical.
If I were to use a buck driver, would that have to be spec’ed as having 6V to the emitter? It seems like a lot (all?) of those specify the input voltage, but don’t specify the otput voltage, where the MT-G2 appears to require Vf of 6v-7v, if I’m reading the datasheet correctly?
I'm not aware of any buck drivers that output a specific voltage, it's entirely dependent on the LED it's connected to. So as long as it's a buck driver, and the input voltage is enough above the LED's Vf to allow proper regulation (needs 1-1.2v headroom), it will work fine.
Similarly a 7135-based driver doesn't determine how much of the input voltage needs to be burned off for a particular LED, it just does it. It's determined by the LED, not the driver.
Drivers control the led by regulating amperage. Ohms law I believe?. If you input a regulated amperage into the led, the voltage will be where it's supposed to be for that amperage. If you go direct drive, then you have no regulation of amperage and if the voltage is high, the led can fry itself because the amperage is not being regulated, so it can run away. I have not tried MT-G2 direct, but I have fried 2 MT-G leds trying to use direct with NiMHs. Both fried even though I thought I had it worked out. You could do it with a resistor, but the problem is which resistor to get it right. How much resistance is in the circuit, etc, etc.
A buck driver that outputs around 3000mA and takes the input voltage you want to use would be the safest bet. Just keep the input voltage +1 volt higher than the 6 volts you want for the MT-G2.
That said, I am going to do one DD with a 6V SLA battery, just to see if it goes poof.
I guess that I’m still of the “I have to see this for myself” type, even if it looks dumb :)…
So when you try that 6v battery (what is it like an OOOLD car battery - you remember those don’t you - use to be 6 volts :)… JUST KIDDING), can you post back, and I’ll do the same with my “experiment”?
6v lead acid batts aren't uncommon, just not used in cars anymore. Stuff like backup power supplies, etc. Much smaller than a car battery.
I'm considering (i.e. really really really want) a L2P MTG2 light, if only I can decide on the power source. There will be compromises due to driver space constraints and cell config, but oh well. Will probably try to keep it around 2A. I just have to decide on 16430, 18350, or something else. Not a lot of good quality cells in that size range, but a necessity due to the voltages required.
Lot’s of spotlights with halogen bulbs run two 6volt lead batteries. Connect the batteries to run in series and find a way to heatsink it and it would be interesting to see what it would do. You might end up with an LED searchlight that can actually compete with an HID and have longer run times without the warm up.
I hooked up a MT-G2 to a LD-29 buck driver (8.4V input voltage, input current 1.2A, output 2.8A, measured on XM-L), it does not even light up the MT-G2 to a moonlight level at any current I tried.
Was that “output 2.8A” at the emitter, and was that a 6V MT-G2? The MT-G2 datasheet seems to indicate that maximum DC forward current is 3000 mA, so if that was a 6V MT-G2, and if that was current at the emitter, you would have been close to the max DC forward current?
The output was on an XM-L, so at the emitter. Yes a 6V MT-G2. Yes you would be close to that 3000mA specified. in my measurements I measured 6.3V is needed for 3000mA to be pulled by the LED.