Direct drive driver?

Flashlight Modding and DIY Parts
1

I'm searching a usable DD-driver that can handle ~10amps. Does it exist?

2

I think I must be missing something. Why not just wire it direct drive, why do you need a driver?

3

I want modes? :P

4

Oh, so direct drive on high? The East-092 may work, but I have never used one over 5 amps. I'm pretty sure E1320 has some extra ones you could PM him and ask. Are you making a SST-90 light or a quad XM-L?

5

Yes, exactly. Like the UF980L..

I dont know yet, I'm currently in the evaluation-process. How is your project going? :D

6

Dry driver is dd on turbo,but I don't think 10a is gonna be possible.

7

Still waiting on the boards, I already have 3 U2s and a 17mm blank. I'll probably order the heatsink from mattaus sometime this week. :D

8

Since this is a project, you could simply buy a MOSFET and ATTiny13A, those are available from ebay. And grab a USB programmer too, it was only $3 last checked.

Some simple uploading (can use drjones program) and soldering and you've got a really kicking direct-drive customizable driver.

9

Yep. That’s the way to do it. Depending on how fast you want the PWM to be though, you might also need a mosfet driver. The ATTINY parts can only push about 20mA out of a port pin, so you won’t be able to turn the mosfet on and off very quickly…

PPtk

10

Assuming a typical 30nC gate capacity, 20mA should be enough for 10kHz without a gate driver, I guess.

There are a couple of FETs easily able to handle 10A, but I wonder about the board…

PilotPTK: What FET would you recommend?

11

The problem isn’t so much the ability of the microcontroller to switch the fet that many times in a second - sure, it can turn it on and off at 10, perhaps even 20KHz… The consideration is the amount of time the FET will spend in its linear region - acting as a resistor. The less current you pump into the gate, the higher the percentage of time that the FET will be linear. A FET operating in its linear region generates a lot of heat when compared to a FET operating at or near RDSon. With only 20mA of current, I wouldn’t dare operate the PWM above 5KHz to begin with. You can try faster if the heat seems manageable.

10A is no problem for a circuit board. Big planes and pours from the wires in to the FET and back out will handle it without issue.

For this application, I’d probably use the Vishay SiR802DP http://www.vishay.com/docs/65671/sir802dp.pdf
Reasons:

  1. It has a max Id of 18.2A at 70 degrees C - plenty of design margin to switch 10
  2. For this size FET, it has a very low gate charge. 15.5nC
  3. Reasonably fast turn-on/turn-off times.
    4) Rated for full turn on at only 2.5V. Most FETS are rated for full turn on at 4.5V or higher (bolded, because this is really important)
  4. RDSon at 2.5V is quite low - only 7.6 milliohms
  5. Assuming zero switching losses (They will be close to zero at low frequency and get higher as frequency increases), only 0.76Watts of heat and 76mV (0.076V) of voltage drop.

Possible concerns with this FET

  1. Only 20V Rated. Absolutely no more than 4 Lithium cells in series
  2. Gate is only 12V Rated. Presuming this will be driven by a uC pin, not an issue.

PPtk