10mm linear driver- discussion thread now

Nobody’s mentioned it yet but there’s no reason you couldn’t do a Zener mod to this and put an mtg-2 in a AAA Minimag. Not much point in it since although you can get the voltage you won’t get enough current from 10440’s to justify it but when has that stopped us.

I’m running 3.01A to an MT-G2 from 2 10250 cells. Mule configuration, 1345.5 Lumens OTF at start up.

Edit: Perhaps I get your point, couldn’t get enough 7135’s on the small board. Shutting up.

Remember the easier you drive an LED, the lower it's Vf. The lower the Vf, the more power the AMC chips have to burn off.

EDIT: I think this is how it works:

With fresh cells (8.4V total) and driving an MT-G2 at 3A, the Vf is approximately 6.2V. This means the AMC chips are burning off (8.4-6.2)x3 or 6.6W of power. Or around 825mW per chip (8 380mA chips).

With fresh cells (8.4V total) and driving an MT-G2 at only 380mA, the Vf is approximately 5.2V. This means thew AMC chip is burning off (8.4-5.2)x0.38 or 1.2W of power.

Does that sound right?

Which kind of brings me back to the question of why are the 350/380 binned 7135 chips the ones primarily used? There are regulation chips with much higher initial current, so why so low that we have to have 8 to get 3A? Or more and more and more…what if we started with a 1A chip? A 2A chip?

It's a number of factors. The primary reasons are because the chips are cheap, small, easily dimmed, and operate down to low voltages. A lot of the higher current chips either can't be PWM'd properly (react too slowly to PWM signals - case in point was the PICColo), are too big, or operate at voltages too high for the applications we like to work with.

Sucks. I'm always looking for new parts but the efficiency of switching regulators is attractive. We just need to get them small enough.

- Matt

The DMG6968U I have used which is smaller than AMC7135 physically gives me ~3.5A on one unit. See this post.
Cheers! :beer:

Mattaus, see my earlier post. I have not tried PWM on DMG6968 but I think it will work.
It’s cheaper/amp too. Let me experiment on that. Cheers!




Looks like I used a good chip.

I have some thoughts about minor changes but I’d like to hear back from the others and have those comments grouped in the thread. Looking good so far though.

Looks great i cant wait :slight_smile: what led did u use to test ?

Pretty sure it’s an XML U2 1A. I bought several from IS on sale awhile back and use them for stuff where tint doesn’t matter. I’m more of a 1D, 3D fan.

Well done, not that l ever had any doubt. Looking good.

Matthaus did all the heavy lifting here. I’m just the beneficiary.


It works. Happy to make changes if required :)


Got mine inbound…can use coupon code for a whopping .38 (10%) discount

Thanks Mattaus. looks to me like another fine collaboration effort.

This isn’t a boost circuit so to run a 2 x AAA host you would need a single 10440 and a spacer.

It’s a regulator right gotcha…I have a few spacers around…it was only $3~ so no biggie

He just used a 13 letter word! :open_mouth:

Nice work Matt, Scott, now get it to 2A and we’ll be delirious! lol JK, unless you have the infamous “Honey I shrunk the kids” ray gun. Scott? Ray gun? Hint hint…

He also has a 1.5 watt version…