A different thought occurred, how sure you are you don't have a wiring error? You could take pictures of your circuit (both sides) and give us a complete parts list. I am especially curious about Q2, MOSFETs do not have a standard in pin outs.

 

A different thought occurred, how sure you are you don't have a wiring error? You could take pictures of your circuit (both sides) and give us a complete parts list. I am especially curious about Q2, MOSFETs do not have a standard in pin outs.

pinout Of Q2, followed
http://circuits.datasheetdir.com/97/VISHAY-IRF510-pinout.jpg
Q1 I followed
http://www.dz3w.com/uploadfile/2010/0807/20100807060727783.gif

Would the surge from a capacitor across R3 burn up the LED? or would it be tolerable with a small cap?

 

Truth to tell, I don't know. It could, and I know 3W LEDs are expensive.

The cap on the gate of Q2 is another story, soft starts are a good thing.

 

I would put the cap across the LED(s).

And Bill made a good point, an NPN bipolar transistor would be a better choice than a FET.

 

Put .1uF capacitor from gate to ground
I got up to 1.86 amps max as opposed to 1.5 max before...
Should I try larger cap?

Voltage R1 R3 Vbe I
7.6 v 1k 0.22 0.42 1.86
7.6 v 1k 0.22 0.42 1.83
7.6 v 1k 0.33 0.55 1.63

There was one time when measuring where the Vbe kept jumping from 0 to .4 volts on the multimeter, the current held steady as a rock tho. (couldnt repeat this error..., maybe bad multimeter connection or the oscillation)

 

What is the gate voltage? If it's at Vcc, then your LED voltage drop is causing the MOSFET to be fully on. Making R3 smaller will indeed cause current to increase, but the NPN will be off, so you will have no regulation. I doubt that a ≈2 Amp LED exists that will drop only ≈2 to 2.5V, which is what you would need with this supply voltage. Try fewer LEDs, or a higher supply voltage.

 

What is the gate voltage? If it's at Vcc, then your LED voltage drop is causing the MOSFET to be fully on. Making R3 smaller will indeed cause current to increase, but the NPN will be off, so you will have no regulation. I doubt that a ≈2 Amp LED exists that will drop only ≈2 to 2.5V, which is what you would need with this supply voltage. Try fewer LEDs, or a higher supply voltage.

No. It should drop around 3-3.3 volts across LED. I was under the impression the voltage source only has to be .7 volts higher than this( for v drop across Rsense)
The problem I find is in practice a smaller R3 does not cause more current. and makes it appear as if Q1 turn on voltage is dropping. Also the voltage to which I read Q1 Vbe is, accurately predicts with the current I measure, via ohms law the sense resistor's actual ohmage.

adding the .1uF cap took the max current I could draw(1.5 amps(with any sense resistor, since Vbe dropped as sense R dropped(in measurment))) to about 1.9 amps max(Vbe dropped, but not as much)

Why I'm a getting a reading on my multimeter that Vbe is so low(i.e. .2v, .3v,.49v..) with different sense resistors?
oscillation of some type?

 

No. It should drop around 3-3.3 volts across LED. I was under the impression the voltage source only has to be .7 volts higher than this( for v drop across Rsense)
The problem I find is in practice a smaller R3 does not cause more current. and makes it appear as if Q1 turn on voltage is dropping. Also the voltage to which I read Q1 Vbe is, accurately predicts with the current I measure, via ohms law the sense resistor's actual ohmage.

adding the .1uF cap took the max current I could draw(1.5 amps(with any sense resistor, since Vbe dropped as sense R dropped(in measurment))) to about 1.9 amps max(Vbe dropped, but not as much)

Why I'm a getting a reading on my multimeter that Vbe is so low(i.e. .2v, .3v,.49v..) with different sense resistors?
oscillation of some type?

Have you measured the gate voltage?

 

Have you measured the gate voltage?

Gate Voltage Current
6 1.56
5.68 0.91
6.03 1.57
5.9 1.31
6 1.57
5.64 0.9
5.68 0.91

*Voltages measured were the same with and without cap from gate to ground

 

Why not take pictures of your breadboard (top and bottom) and post them?

I'm thinking of wiring my own for experimentation. I suspect my setup is a bit better, unless you have an oscope you haven't mentioned?

I would probably use a protoboard like I showed earlier.

Suggestion, replace the LEDs with a current meter or a 5Ω resistor and measure the current through this setup.

Another thought suggested itself. MOSFETs are ESD sensitive (very). It is possible you have a bad component.

 

Why not take pictures of your breadboard (top and bottom) and post them?

I'm thinking of wiring my own for experimentation. I suspect my setup is a bit better, unless you have an oscope you haven't mentioned?

I would probably use a protoboard like I showed earlier.

Suggestion, replace the LEDs with a current meter or a 5Ω resistor and measure the current through this setup.

Another thought suggested itself. MOSFETs are ESD sensitive (very). It is possible you have a bad component.

No oscope hah
I did replace the LEDs with an ammeter. //no change

I initially tested the circuit on a bredboard with 2 mosfets(irf510 ad irfs630) They worked at low currents and I went along with the build with the irf510 which had a nice metal backing. I have easily touched it with all sorts of things while building...
Perhaps I should solder in another to test?

 

No oscope hah
I did replace the LEDs with an ammeter. //no change

I initially tested the circuit on a bredboard with 2 mosfets(irf510 ad irfs630) They worked at low currents and I went along with the build with the irf510 which had a nice metal backing. I have easily touched it with all sorts of things while building...
Perhaps I should solder in another to test?

Don't you have a voltmeter? You don't need a scope to measure DC.

 

soldered in new mosfet... was careful not to touch pins and such.
current went up to 2.3 amps!, (near where I want it)
Then I soldered in a resistor I had clip-leaded on and moved some stuff around, next time I tested it the current didn't go up much past 1.3 amps, not sure if I didnt solder resistor on well or what, but Im off to slumber while the batteries charge. will test later.

 

\(R__{DS__{(on)}}\) will be 0.54 ohms for IRF510 when \(V__{GS}\) is atleast 10V i.e.. to turn on the mosfet you need 10V at its gate ,your supply voltage is 7.6 increase it.

Good Luck

 

\(R__{DS__{(on)}}\) will be 0.54 ohms for IRF510 when \(V__{GS}\) is atleast 10V i.e.. to turn on the mosfet you need 10V at its gate ,your supply voltage is 7.6 increase it.

Good Luck

This is an analog circuit. It doesn't require the MOSFET to be fully on. Feedback through the NPN forces Vgs to the value that allows the MOSFET to pass current equal to Vbe/R3. At least, that's the theory.

 

Tested new mosfet, It appears to be working as it did before.
Current only goes to 1.67 amps.
It appeared to have worked at first, but now its back to where it was...

 

What is the part number of your LED?
How many do you have in series?

 

What is the part number of your LED?
How many do you have in series?

Just 1 T6 led to power. Would like 2.5 amps on 'high' mode. I have switch that I have that puts more resistors in paralell.
I have tested with the led and just an ammeter from vin to drain. I will test with just the ammeter with current setup again.

 

Just 1 T6 led to power. Would like 2.5 amps on 'high' mode. I have switch that I have that puts more resistors in paralell.
I have tested with the led and just an ammeter from vin to drain. I will test with just the ammeter with current setup again.

I have been assuming you had 3 LEDs in series, as in your schematic. Perhaps you mentioned previously that you had only one. If not - your bad!

 

I have been assuming you had 3 LEDs in series, as in your schematic. Perhaps you mentioned previously that you had only one. If not - your bad!

sorry, I dont know if I did.
Any idea though, why its not working right?