Buck and Boost Drivers, Testing, Modding, and Discussion (Pic Heavy)

solRNY, protection circuits are an additional hurdle which reduces a cell's discharge performance. Your protected KeepPowers are more than likely low drain cells with standard discharge protection circuitries, causing additional output voltage sag. The equivalent series resistance in one of those protected batteries is equal to the internal resistance of its low drain cell (high, above 50-60mΩ) plus the resistance of the probably two 8205A MOSFETs (close to 15mΩ) which switch the cell's cathode. This is very high and completely inadequate for even remotely close to high drain applications.

A high drain cell's internal resistance is generally well below 30mΩ. Mooch, the vaper's community reviewer, concluded that low drain cells are only worth it at up to half of its rated maximum discharge current (or so) versus the 2800-3000mAh high drainers. It can be argued that the test's cutoff value is very high (3.2V) and I agree, but anyway it serves as an enlightening reference: Which High Capacity 18650 is the Best? -- Shootout Bench Test Results @ E-CigaretteForum

V1 posted on Thu, 08/10/2017 - 15:17; redundancy typo fixed.

thanks for the battery education. I will use high drain unprotected cells.

0K my technical fellows, just found this lost beauty (I hope): CF FX-17A 17mm 3V - 9V 2.2A 1 cell or 2 cells 2 Groups 3 to 5-Mode Driver Circuit Board @ Kaidomain

No relevant information found for the driver. My observations:

  • Reverse polarity protected. As Jensen567 explained in this same thread some messages ago for the H1-A, this means the driver innards' ground plane is isolated from the host, which is heat transfer detrimental.
  • Amount of modes and spacing: 5-40-100%, with optional strobe and SOS. :THUMBS-UP:
  • Last used mode memory, 3 seconds. :THUMBS-UP:
  • Stock output is 2.2A, sense resistor :facepalm: 0.1Ω , 0.22V sense voltage. This is inefficient, a lot of power is lost here.
  • Height is listed as 8mm, spring excluded I believe.


It's 17mm, but wrong in every other aspect. An LD-29 beats it hands down.


I have a couple of those on the way already actually. Im guessing it is QX9920 based but will find out. Plan to use them in some converted mini-mag triples, so not going to push much current.

Mmmkay fellows, finally found and opted for this “host” of my liking: ThorFire TA13 (Sofirn SF30?) questions

Excellent driver boot in there, doesn't it? It probably is the better zoomie flashlight for modding one can buy in that price range. ;-)

Not gonna mod it straight yet, looks nice, no hurry. I'll buy parts as I can slip €urobucks for it, economy is tight now so we'll see.


I have a Sofirn SF30A on the way from Aliexpress because of this tread on TLF, I have a “spare” driver from a Jaxman X1S I plan to use with a XHP35 HD.

Thanks for the link, khas. I'd have preferred for those %@#@&$ not to “hide” the pictures.

Something is said about the led sitting 5mm “too low”, or the lens not being able to get close enough above the emitter, limiting flood. Don't know how do they solve that.

Mine isn't going to be fully modded “straight away”. As it comes from the warehouse I'll just put a dedomed emitter in it.

Do XHP35 HDs dedome well?

They probably dedome as well as XPL HD. Why not just go with XHP35 HI though?

I'll probably go with an XHP35 in a near future, for now I'll just swap the stock XM-L2 with a dedomed XM-L2/XM-L, the stock driver isn't meant to handle high voltage emitters.

The XM-L I have is an old school emitter, produced maybe by the end of 2013, this means it can probably take some high current without blowing up.

However, I still have an old XHP70 M4 40H bought from Kaidomain… (still available on their “AliExpress branch”: https://www.aliexpress.com/store/product/Cree-XHP70-M4-40H-4000K-Neutral-White-LED-1-Piece/708513_32683288528.html). Do you believe, Jensen567, we could reconfigure the H2-C output voltage range to start off from ≈5.xV? Should be feasible, isn't it?


Yes changing the output voltage on H2-C should be possible, that’s my next endeavor as I will probably need to do that for my scratch build anyway.

I see the H2-C manages ≈2.5A in turbo at 12+V with ease and efficiency, at least as long as the input is 2S. How about stacking an R050 (or two R100s) in parallel with the stock ones? That would mean 3A for turbo, 2A after 60s stepdown. I'd just need a Sinkpad XHP70 6/12V MCPCB.


Cheers ^:)

The thing I dislike about 12V XHP50/70 is that on all the DTP boards I have seen, the thermal pad is no longer isolated, and sits at the 6V midpoint of the series chain.

Most lights have the battery negative connected to the flashlight body, so if the thermal pad is not isolated, and the board is DTP, we need to worry about isolating the MCPCB from the flashlight body which can be a pain.

It can be alright if you use good isolated thermal pads (expensive), but won’t ever be as good as proper thermal paste or soldering. That’s why I don’t like running the XHP50 or XHP70 at 12V. Yes the problem of isolation can be solved, but there are very few benefits of running in 12V mode for flasglights so it makes it not worth it for me to worry about the isolation.

I was to get one, but finally went the XHP35 way. In case anyone is interested: Sinkpad XHP70 21mm Copper 12V or 6V @ LEDDNA

Ordered from LEDDNA once, the shipment took more than a couple months to arrive home.

By the way, ordered an H2-C, an XHP35 HI C2 3000K CRI85+ (according to datasheet), and a 4-pack of high drain Aspire 18350 1100mAh cells. For the ThorFire TA13, right. Gonna be a handsome aspherical thrower.

I am to mount an additional R100 on the sense resistor stack. This means 2.25A on turbo stepdown to 1.5A. How much effective throw is to be expected from it fellows?

What is the effective diameter of the lens?

Luminance measurements of an XHP-35 HI LED can be found here.

You can use the 12v Sink Pad no isolation needed! :+1:

Wasn’t aware a board like that existed. Adds some extra series resistance but probably not too bad. Thanks!

So I somehow managed to break the UI on an H1-A today. One of the new ones I got yesterday from KD. I put it in an M2 quad host running 2S2P XPG3s.

Funny part is that this is the least modded driver I have, all I did to it was bypass the spring with solder braid, replace the stock output wires for 22AWG, and file the edge of the ground ring a little bit. Stock sense resistor and stock output voltage.

The boost converter is still working since the light turns on. I just only have one mode, which seems to be maybe 40% - 60% power. Definitely not full power as the light can run for a long time without getting hot.

I’ll pull it apart when I get a chance to investigate, for now it is at least a useable light.

How is that copper braid bypass working for you?

I kept burning through it on my L6.