comfychair-inspired quick+cheap+lazy single-sided 17DD FET-driver (poor man's nanjg92)

.Yes that line is correct, I soldered the FET-drain to the pad where the 7135 used to be.
.

.
Thanks, Iā€™ve been looking for a cheap and easy Fet+1 driver for ages. Cheers but im no programmer I just copy what the clever people on here write and alter it until it works!

My driver mod shows only 3,3A at the tailcap. Did there something went wrong? I have tested with different cells.

I soldered all like shown in the first post. I soldered the LED - to the top of the fetā€¦the old minus isnt soldered to the fet

What batteries have you use? Are they fully charged? What is the gauge of your DMM probes? Did you do spring-mod? What LED are you running with?

Pictures really help. What FET are you using? Many wonā€™t turn on fully at battery voltage and can limit current.

Thanks! The Led is a Cree XM-L2. I use fully loaded LG MH1 Cellsā€¦.there shouldā€™nt be any problem.

I used these Mosfets from MTN: http://www.mtnelectronics.com/index.php?route=product/product&path=25_76&product_id=439

Iā€™m at holidayā€¦so I haveā€™nt any pictures actually :slight_smile: But I soldered it like the first post!

What are DMM probes? Spring isā€™nt modded right now.

Robert

An additional spring bypass should help a lot especially in a 1 cell configuration.
If you are using standard DMM probes, they are probably too thin, you should use short and thick ones (Messleitungen) :wink:

Yeah if you are using stock DMM probes & have not bypassed the springs, that is probably it. Bypass the spring & use fat short wires from the DMM & i am sure you will get higher amps.

Which dmm probes you prefer? I will do the Bypass as fast as possible.

Just try some short fat electrical wire you have, almost anything shorter & fatter than the normal thin DMM wires is better for measuring high amps.

DMM, digital multimeter. The probes are the metal bits at the end of the wires that plug into the meter. LEDs are voltage sensitive and maintaining high current through them extremely dependent on eliminating voltage drops in the circuit which the meter is now part of. Other voltage drops occur in the switch and springs. Minimizing all 3 should result in a dramatic measured difference.

Very well explained. Thank you very much.

So I should use the wires without probes? And replace them against fatter and shorter ones. But how the wires fits into the meter without probes?

And whatā€™s with the dmm wires at the led. You soulder them for measuring?

Use banana plug, just solder your 22awg (or thicker) wires onto the banana plug and you have a pair of good test leads.

I even have solid copper clamps soldered to my (short, thick) DMM-leads, because I found a considerable voltage loss over the iron clamps I used before. I could not find solid copper banana plugs btw.

Congrats for djozz and Major.
Just ordered the banggood driver and eBay fet. As soon as they arrive wil try this in my Uniquefire 1504 host that I just received.
I really hope that this thread receives much more attention and participation from BLF members. This mod has a fantastic cost benefit ratio.
Keep the good work djozz/Majorā€¦.

I did that with a 105c today. ā€¦
stripped the 8 7135 and mounted the fet (taken from a computer Mainboard )

low and high work bit mid does notā€¦
could it be that a fet has problems with a specific pwm ratio?

amps are nearly DD :wink:

I know not enough about FET's to know what happened to your med mode, I just know that there are several types, some work as a discrete on/off switch, others allow intermediate currents (like the FET on the LD1 and LD2).

thanks mozart. Note that in post #10 there's a link to an even cheaper ak-47 driver (set of 2 for $2.89) , from dx. Dx does not mention the type but they are genuine ak-47A drivers.

edit: dx is sold out at the moment :-(

A common issue is that output in each mode is a fixed percentage of maximum so when you up the maximum by replacing the 7135ā€™s with an FET the middle and low modes get boosted as well so low mode might be brighter than one might wish. This is normally corrected by adjusting the firmware to lower percentages.

Another is that if the input capacitor is in the wrong place voltage spikes can cause malfunctions.

I have bypassed the springs at switch and driver. Know the xm-l get burned :frowning: I think thats to much current. Is that possible? It is hard to measure without destroying another led.

LED mounted on DTP board, & attached to a light/heatsink, you should not be burning them.