I just pulled out the unit I used for some testing earlier in this thread. On my sample (which has been used some for testing on the bench) it’s ~11mm overall (including PCB) without compressing the driver. Compressing it brings it down to about 10mm overall.
The soldering job holding the toroid in place is not perfect. It does not look like I ever reworked those solder joints, I think it’s safe to say it’s the factory job. That’s where the extra height on the one I have comes from.
The PCB is 1.4mm thick.
The toroid is about 5.5mm thick, including the wrappings.
The tallest component other than the inductor is the Schottky diode, it’s about 2.4mm tall.
I think if you were very careful in reworking this board you could get the total height down to 9.5mm or 9.4mm. So in a best case scenario the height above the board could be reduced to 8mm… maybe.
I actually planned to put this LD-40 in a Dereelight Xsearcher of my friend, but the driver compartment of its pill is very small and shallow (only allows ≤5mm of driver height).
Looks like it is a no-go. Well, is there any other 16-17mm buck driver with shorter height? Or should I go with the zener modded Nanjg 105C driver?
Since I am going to run an XP-E2 with two batteries, using a zener-modded Nanjg 105C will give me pathetic efficiency, right?
How much current do you want to use? For 2.0A to 2.5A you could build a Knucklehead v3.0. See here: Mattaus - Knucklehead Buck Driver Discussion Thread Mattaus has removed the OSH Park listing because of the driver’s inability to handle his 4A design goal reliably (at least unpotted), but RMM has tested the driver to be fine at lower currents like 2A. I’d send Mattaus a PM, but if you can’t get in touch with him lots of us have the appropriate files, myself included. If you want to order boards you’ll definitely be able to.
How much space do you really have? The commercially available ‘thin’ options I know of are the two boards from LDCH, LD2C and LD29. The LD29 is ~18mm or ~19mm but uses a 20mm or 21mm contact board. It’s also still taller than 5mm. Also both drivers provide a little too much current probably without a resistor mod for a lower current.
Anyone know if this’ll fit in the BLF X6? I’d like to be able to use primaries and iirc one of the components of the stock BLF SE driver tops out at about 5V in.
Yes about 2.5A is all I need for my friend’s Xsearcher. But building a driver is out of my capability at the moment - I can start learning, finding sources to acquire the necessary parts and put them altogether, but that will take very long time. Thanks for the suggestion, it really looks exactly what I need but I am hoping to go for something which is readily available and not too expensive. This driver I am looking for is for my friend after all, not for myself.
Yeah I have thought about the LD29 too… but like you said it is too big in diameter. Maybe I can remove the 21mm contact board but I doubt if the LD29 itself can be ground down into 17mm without fully removing the GND edge.
The Xsearcher driver compartment only takes 17mm in diameter and not more than 5mm in height (most probably 4.8-4.9mm).
One more question to you Alex: Currently my friend’s Xsearcher is using the zener modded Nanjg 105C to run an XP-E2 (3.0V) with two batteries (8.4V) and after 2 minutes operating on high the light starts to flicker terribly. Can that be due to the overheating of the driver (7135 chips?).
Thanks a lot.
EDIT: Found a similar discussion in this thread . But I still wonder if the flicker is caused by the overheating of the 7135 (or the other components).
Hah. Attempting to drive an XP-E2 with a Zener-modded Nanjg on 2s is not something I’d personally bother with. You’re talking about dissipating around 1.4W per 7135, or about 11.2W total! That’s pretty crazy. I’m impressed that the 7135s survive at all. The flicker is almost certainly because the 7135s are unhappy. At 1.4W dissipation each, even with really good heatsinking I don’t know if they can become happy. There’s just no fixing the situation. You’re talking about <50% efficiency. This can only lead to bad things. Did you ask “Or should I go with the zener modded Nanjg 105C driver?” … after already doing it?
As far as fitting an LD29: I do not think that any material can be removed from the LD29. I don’t think there’s a cheap fix for this. Even modding an LD40 with an SMD inductor will only reduce the height by a little.
Your best option other than the Knucklehead v3.0 may be purchasing a replacement 2MT-S assembly from Dereelight.
What happened to the stock driver?
If I was going to run a Nanjg driver I’d just put in a dummy cell and run it 1s1p.
Haha yes actually this has already happened in my friend’s Xsearcher, but it was not done by me. My friend has his Xsearcher modded by an oversea modder and the modder put in a zener modded Nanjg 105C driver. When my friend ran it with two 18650 after more than 2 minutes the light started to flicker badly, note that the emitter is a dedomed XP-E2. Last week he sent the pill assembly to me and see if I can solve this flickering issue and change out the driver for him.
I don’t feel like to reveal who is the modder here thus I asked about this as a ‘if’ question earlier… and now you know the whole story.
As for what happened to the stock driver… the modder kept the stock driver (did not return it) without my friend’s consent. I wish my friend has the Xsearcher stock driver with him now so that we can probably solve this problem much more easily. AFAIK the 17mm Dereelight buck driver is quite okay to handle high current. Maybe stacking another resistor will do the trick.
Wow. I'm surprised that it is even working for 2 minutes. I have a few very flat 17mm buck drivers that I got to try out, but they're only around 1.3-1.5A and have stupid modes (strobe-sos).
Do you mean the combination of XP-E2 + zener-modded Nanjg 105C + 2*18650?
Luckily my friend just used the modded Xsearcher for fun and has never operated it for long-continuous time. I have disassembled the pill assembly already and now I am going to put in a new Nanjg 105C with stacked chips and running a dedomed XP-G2. This way my friend is always going to run it with a single 18650 only.
Yeah this is better apparently, but is this FET similar to comfychair’s one which is capable of high current (low internal resistance)?
And I don’t even have a SOIC8 clip yet so… Anyway I have already added one SOIC8 clip into my cart in FT and ready to push myself to learn it later lol.
Sorry to bump this thread, but i was wondering, if i can use this driver to drive an MT-G2 using a single cell or to be precise a 3PS1, i still don’t understand how this could happen if this is a buck driver and MT-G2 needs 6V.
I’m planning to mod a Courui D01, oh well i could also buy a 6x 16340 IMR, but I’m just afraid of all those batt in series.
Since it’s mentioned in the IOS page it can drive MT-G2 or 2 led in series…Can someone shed some light on how this could happen
It can't drive an MT-G2 from a single cell. For that, you would need a boost converter (one in 17mm with any sort of significant current output doesn't currently exist).
You can drive an MT-G2 with this driver, as long as you have 2S or 3S cells on the input. The MT-G2 has a forward voltage of around 6.1V-6.2V at 2.4A, so even with 2S cells you have some pretty good voltage overhead.
I will go ahead and chime in on this since quite recently I have had a fair deal of experience with this driver.
The short answer is NO you cannot drive an MT-G2 with this board and only 3.7V (one cells).
Typically using 2 cells would have worked. I have been using it to power XHP-50s on two cells with an added sense resistor at 3.7A and it worked well.
However I will say this. The newest batch of these I ordered from IOS have some issues. The old version could drive an emitter at an almost linear rate. IE you would get full power to a 6 V emitter off 2 cells. This new batch it would drive the emitter at only around 600ma :( Unless you use higher voltage. It would take at least 3 cells in series to get the full amperage to the emitter now.
FOR SURE the inductor on the new batch is smaller, and I suspect other components may be changed as well causing this huge loss of performance.
I have sent one of each to someone with FAR more knowledge of this stuff than myself to do some testing, to find out which components will need to be replaced in order to get them working like they were before.
Just saying: This driver powers an MT-G2 at an ok level. Light is a Convoy C8 with a widened, smooth reflector, aluminium pill and 18650 extension tube.
.
To my eyes, output is pretty much the same as a 3.0A driven T6,XML2. High current draw with two freshly recharged laptop pulls is 1.44A, med 540mA and low 50mA. Should’t 2.4A to the MTG2 result in a higher output than an XML at 3A?
.
With two partially depleted LiFePO4 cells on high mode, there was some weird flickering/dimming that looked like a bad connection but didn’t respond to knocking the light and disappeared in med and low modes. Current draw on high was 1.70A.
.
With two samsung 30A (or 30Bs) charged to 4.30V, current draw on high was 1.55A. I thought this was weird since the driver is supposed to be buck and as voltage goes up, current draw was supposed to reduce…
.
Either way, with the two laptop pulls in the light, I added on a third cell to see what happens. (driver is supposed to handle up to 16V.) BRIGHT, 2.45A Yay!! oops, battery slipped, reset, nothing… battery didn’t slip, something died. I assume the driver died and not MTG2 but it was not the usual stinky, dead IC smell so can’t be sure yet. ;/ Will confirm this evening when I get home.
EDIT: after removing the driver, i got the familiar, burnt electronics, stinky smell. using a 40x magnifier, it looked like the sense resistor alone was burnt. I didn’t have anything similar so swapped in a R015 resistor from a laptop pack… Nothing happened ;( I can’t see anything else that is burn’t.
.
Zener modded a Qlite and used that as the driver instead. (I’m surprised to see that moonlight with the MT-G2 is the same brightness as an XML2… )
(I’m surprised to see that moonlight with the MT-G2 is the same brightness as an XML2… )