Convoy 17/22 mm SST40 Driver temperature protection removal

Hi BLF-ers

Simon is selling this driver:

I have it in my L21a and was wondering if it is possible to address the NTC or PTC or whatever temp measurement component is used.

I like the 4 mode driver, but the temperature management is set to 55 degC, and i would like to bypass it or adjust the setting. (Or remove the component if possible)

Does anyone have experience with this driver and doing so?


Thanks Simon ! (For those not familiar, he is the owner from
See posts below how to remove the temperature stepdown / temperature protection of the SST40 drivers which Simon sells

Regards P.

I bought the FET version of this today in 17mm since it runs the LED harder than this version can. I don’t know if it has a eprom on the UI like the one I got. It might be built into the ic.

So, i just asked Simon [the official owner of Convoy @] and he answered me with the following images in This thread

So tonight i had to find out whether it was actually a PTC or NTC, to remove the item or make a solder connection between the pads.
To find out i measured the component Simon pointed out at room temperature which gave me around 90K, after warming up the driver with my wife’s blow dryer the resistance dropped to 35K.

So i desoldered the NTC and did a second second test on only the component (which gave different readings ofcourse, since measuring resistance in a loop is actually a no-go)
No load resistance gave me 105K and after heating up with the blow dryer gave me around 35K (it raised quite fast, that’s why the picture shows ~60K)

Now the light gets hot after a few minutes, which means, mission accomplished, now i have an L21A without temperature protection!

Removed the NTC

Good job Yokiamy, i am interested in this mod also but i cannot see the pictures in the Convoy tread. Is it possible to reupload them to imgur, please?

Somehow the images don’t showup on imgur, if you exchange email by PM, i can send them to you. Maybe you can try to share them after you received it

Sure, PM sent.


Got the pictures, if anyone has the problem of not seeing them in the Convoy tread, here they are:

To be honest i did not think that it would be as easy as removing single NTC resistor, gotta fix my C8 now. :slight_smile:

Me neither, but in Dutch we have a saying which may not sound as good in English, but it’s something like:
Not tried is always wrong.

So i just tried to ask Simon, and he was willing to help (and it was actually dead easy)

“You miss all the shots you don’t take” would be a pretty good alternative :slight_smile:

That’s even better!
In Dutch we say, nooit geschoten is altijd mis!

Thanks for this, very useful info! :beer:

Just buy the Ramping driver, it’s much better

can you give a substantiation?

Now it just sounds like a personal preference, without explanation

Interesting, understandable, logical. The temperature protection gets activated by the resistance reduction in the NTC. If the NTC is removed the resistance “gets high” and cannot change. The temperature limit is thus removed.

Concerning the ramping driver, I modified a ∅22mm unit for a little bit more current (8.8A theoretical, 8.1+A measured) and two ∅17mm units for less current (5A and 4A, hitting ≈4.75A and 3.8+A when measured). The ∅17mm unit set for 4A had a small problem, with the first seconds of the ramping unseen and the 3rd mode setting clicks also unseen. Couldn't investigate the issue, the flashlight was for a colleague and he came home, got it and went away despite the thing.

The ramping driver has a weird UI which may not be everyone's liking.

Can you explain how you made the ramping driver less current? Would it be possible to run at 3A for an XML2? If you had to swap out chips then I won’t be doing it

I like the 6A driver but wish I could make it ~7.5 - 8A :confused:

By swapping out the sense resistor stack. Comes stock with 5mΩ, this means 40mV theoretical (I = R × V) sense voltage, although seems to be a bit less in practice (?). I added an R050 on top of the stock R005 in the “stock 8A” ∅22mm driver which ended up measuring 8.1+A. Concerning the smaller drivers I installed two stacked R020s in the one I set for “4A”, and for the one I set at “5A” I went with an R010 plus R050 plus R200.

For 3A effective a couple R025s may do.

Wed, 08/21/2019 - 11:43

1 Thank

@Barkuti what do you think if I just bypass “5mΩ” with the wire to run XHP50.2 3V I’m not sure if it a good ideal.

Barkuti, which component is the sense resistor stack? I think I see the R005 on the ramping driver, but do you know which it is on the 6A driver? I assume R010?

So what I’m understanding is that more resistance = higher amperage… Now to work out the right number to hit ~7.5A :slight_smile:

Edit: I’ve found a similar driver to the 6A one, but this is apparently 7A: - looks like it uses R015, but there’s a few other components of different values, too.

No, by adding the same resistor on top, your total resistance will be half of the value.
(Electronics series parallel basics) This will increase current output, many people did this with the L6 resistor mod (R100 or R120)

If you do that the driver won't be able to regulate at all. This means no ramping and no modes. I haven't tried it, though.

You still need to tweak the sense resistor stack value for it to work. You could try an R004 on top, it could work well or it may not. Add some heatsinking to the thing in such a case, as you'll be above the maximum power rating for single 1206 resistors (a couple stacked one doesn't changes this significantly). A good mound of thermal glue above sense resistor and MOSFET could help.

If the driver uses R010 for 6A this means sense voltage is Vsense = R × I = 10mΩ × 6A = 60mV. If you want to add 1.5A, you need Rextra = V / I = 60mV / 1.5A = 40mΩ or an R040 in parallel with the existing R010.

I have not measured how that driver behaves actually but if you really want to be sure to hit 7.5A, add some more conductance is my advice. I'd look for an R030, which would raise the maximum current to 8A theorerical (6A + 2A), albeit it would leave the actual figure in close proximity to 7.5A effective due to MOSFET heating, imho.