H17F - programmable driver with full thermal regulation

Spark?

Adobe :stuck_out_tongue:

Now that was quick! It's PostScript, and chadvone has won :) Gratz!

Thank you very much… I am looking forward to giving this a try.

I was 99% sure I had the right answer on 1-11-2016, right after I gave a wrong answer.

It has not been quick for me.

Red will get a long walk tonight.

Congrats chadvone!

Congrats on the win, chadvone! :party:

Thanks, my first win.

Finally, congratz chadvone! :party:

There is a lot I don't know about PostScript!

Nice going, chadvone.

Congratulations chadvone. Never would have guessed it.
On the other hand: everything that happened after I learned how to put the ram in the rama lama ding dong is one big blur.

Which mosfet are you using DrJones ?

Congrats chadvone.

DMG6968U. 10A seems to be the practical limit for the driver under good conditions (Nichia219C triple; while a driver with a big FET gave about 15A, as far as I remember), and with an XM-L2 it gives 5.8A, which I find quite sufficient.

Received my H17F driver today. Was not expecting it this soon. Complete surprise. Actually made my day after my order from RMM didn’t make it.

Low is so low you can see the detail in the LED.
High is bright. Much brighter than Stock 3 speed.

UI Every one has there idea of a perfect UI. This is my perfect clicky UI. You can make it as complex or as simple as you would like. Brightness adjustment, there is brightness level for anything I can think of.

I was interested in this driver since I first read about it.

Great Job DrJones. All my click switch lights need these.

Chad

I got mine about a week ago and I’m absolutely in love. I’ve been calling it the ‘dream driver’ as it’s got everything.

One question, on the battery check feature, mine seems to blink 12 times on a full 4.2v battery. I thought mine was getting relatively dim on turbo, but the battery check still blinked 5 times, and it read about 3.2v on the DMM when I checked it.

The manual says each blink is 0.12v above 3v baseline, which is 10 blinks for a full 4.2v, and should have been only 1-2 blinks for my drained battery.

Was this revised in a newer version of the FW?

From the manual

The number of blinks indicates battery voltage, about one blink for every 0.12V above 3.0V without load (~10 is full (4.2V), ~4 is pretty empty).

I am sure there is some variation in accuracy. 4 is pretty empty.

I think mine sometimes blinks 11 times

Wow, this seems like a great driver with an impressive set of features. I just ordered 2; one for my s2+ with single XP-L and one for a single cell 26650 triple XP-L that I’m planning in a trustfire a8.

I gathered from this thread that the FET used has more on state resistance than, say, the regular FET drivers from mtn electronics? If this is the case I might not want to use this for my triple if I want the possibility of maximum current.

I’m interested in the thermal control. Is it important to sink the driver with something thermally close to the emitter? The PID control sounds cool. The driver can access all 24 brightness modes to control the temperature?

Is semi turbo something like 50% PWM of the full turbo?

Thanks.

I measured my triple XP-L at 10.5A on the MTN-17DDm driver with 18awg wire, and 10.4A on the H17F driver with 22awg wire. I was a little concerned about the “loss” of output, but for me the practicality and features of this driver heavily outweighs any losses. I don’t think a difference in output would be discernible outside of very controlled testing.

I just have my driver in the pill of my S2 like so:

and I ran it for over an hour one day on turbo, and the thermal regulation did its thing. There was no controlled testing, but it seemed to lose approximately half the brightness as it throttled output. I couldn’t ask for more.

Moonlight is super low. “Moonlight” on the S30 (1 lumen) vs moonlight on my triple H17F:

The only thing I would change is the spacing between the lowest two modes. I would lower the output on the second level, personally (if possible). It’s quite a jump in pitch darkness.

Sorry if this is a dumb question, but will this driver be suitable for this particular emitter?

http://www.mtnelectronics.com/index.php?route=product/product&product_id=572

Yes, that’s what’s in my light posted above.

I’m glad you like it :slight_smile:

PrinceValorum: The voltage measurement depends on an internal voltage reference in the MCU, which is known to vary a bit - in uncommon cases up to 10% - you probably got one of these :frowning: Can’t do much about it. Calibrating each MCU would mean too much hassle (and increase the price…), and there’s no space for a better external voltage reference on the board.
Lowest two modes: These two modes are the lowest possible PWM modes for the single AMC7135. The AMC7135 takes about 2µs to switch on, and the lowest mode sort of switches it off again right then. The second mode already gives the AMC time enough to really switch on, that’s why that step is quite big.

EasyB: Yes, the FET has a higher resistance, because it had to be smaller. But the driver is good for 10+ Ampere, as several people meanwhile measured (one even 11.5A).
Thermal regulation works smoothly, i.e. does not jump in steps of those levels. The algorithm seems to work well under a wide range of thermal conditions and setups, so I’d say it’s not important to improve thermal contact of the MCU to the LED. If the control behaves oddly, you could try that though.
Semi-turbo should be halfway between 100% and turbo - though that might vary a bit due to non-linearities in direct drive.