Exactly…a FET is more or less a “solid state switch” when the gate is activated source and drain are connected like a wire the resistance of that wire is the mOhm rating from the spec sheet…the more resistance and more current = heat generated by the chip…but it will push whatever voltage that is on the source thru to the drain, the amc7135’s are regulators the FET is a on of wire (not to insult anyones intelligence)
This isn’t particularly for you…but I too learned alot by watching this video
Again, not to insult anyones intelligence…but to educate {I too was mistaken on what the FET was and how it worked compared to the AMC7135}
Take pictures. Basically the sense resistors must be in the path of the current (of course you’ve already said that these are) and the buck IC must be able to see both sides: in other words if one side is GND (we assume the buck controller is on the same GND) then the other, non-GND side of the sense resistors must have a trace going to the buck IC (preferably a short trace). The buck IC needs to be able to measure voltage drop across the sense resistors.
That is my understanding too. Which means the resistors are not limiting current on their own as they could not possibly handle 11 amps of current and they would not appreciably lower voltage in a drive that can buck from 16V down to 3V. The they must be helping the driver “sense” voltage differential for controling current. It’s just not happening directly at the resistor bank. Maybe through the FET’s drain pin.
EDIT: I was replying to WarHawk. I will take some pictures and post tomorrow. Thanks guys.
Try it, it’s already broken, what’s the worst that can happen? Probably try the 70n06 first. Note the 70N02 was chosen for its extremely low RDS (on), its one of the best performing (lowest resistance) FET’s out there, so don’t expect the light to perform any better than stock from this component swap.
I put 70N06 but the performance is very bad (is better then with 35N03 defective)
what can I do? can I change the resistor? what value?
or I can buy 70N02? what about AOD510?
It depends. You're working with a blue SRK driver, like the one in the first post? Still using the stock MCU or are you using another board with an attiny13 as the controller?
If it works without the gate resistor then it's not needed. If it doesn't, it is.
Why does it need to be removed? Does it work better without the gate resistor? 10 ohm is not enough to have a negative impact on the FET's switching speed, if it works just leave it there and move on to something else that matters.
The drivers in this thread are single cell (3-4.2v) direct drive, as in SkyRay King and the like. I have no idea what driver you have there and what the power source is, if it's able to power 7 LEDs in series.
Helo
I have the same torch as skyray king but the name ultrafire. It has 7 Xml t6 leds in series, driver blue like in the first pages of the forum,
batteries 4x 18650 paralel so 3.7 - 4.2v
When I turned the torch on it stopped working after few minutes. I found FET 35n03 defective so I removed it and put 70n06 ( because I dont have any other at home). Now the light is very low. What to do, to incerease the light power ( the current to the led emitors)
Your LEDs are not in series then, they are in parallel. It would take around 23 volts to light up 7 in series, and this is not a boost driver.
Link to the datasheet for your exact FET, please. The same markings used on parts from different manufacturers are not the same, they can have wildly different specs. Make sure you have the right manufacturer, the right part number, and the right revision (if there are any markings on the chip to indicate that).
How did you determine the original FET was defective?
I'm not going to give any specifics for what to check where to diagnose the low output thing until I'm positive we can get on the same page about the actual parts being used here.
Ok…I will open the torch and look for the exact Fet marking
Vith original fet the torch was not working so I connected oscyloscope to the gate and I saw that the imput signal to the fet is OK ( i saw 2 modes, one mode full and one mode 200us ON and 100us OFF - low mode)
Then I changed the fet with the 70n06 and the torch started to work but very low power…
If your '70N06' is by Unisonic, the first thing that jumps out at me is that the gate threshold voltage is very very high, compared to the parts known to work well in drivers like this.
