I have a wire with current I1 going thru it. As the voltage and I1 rise there is a transistor Q which turns on and it in turn stops the current from flowing. I1 and the voltage referenced is tied to the base of the transistor.
My question is:
When I1=.90amps Vbe = .63 volts
When I1=1.3 amps Vbe = .58 volts
When I1=1.5 amps Vbe=.38 volts

Why does the turn on voltage of the transistor keep dropping as the current increases?
(q1 is 2N3904)

Added picture upon request. Making it a bit more specific to my circuit

Ok ,
Now that you posted a circuit schematic,
From what you wrote and the schematic given,it looks like you using R3 as a current sense resistor, and the mosfet as a voltage controled current source, so if your current increases too much, you want the circuit to fold back and stop the current, this set up will not stop the current from flowing, as you noted in your description.
This circuit is acting as a feedback voltage regulator.
The reason the voltage is dropping at the base with increased current flow, is because Q1 is conducting at its Vbe value of 0.7v. , but what is happening as an increase of the voltage across R3 is sensed at the base, Q1 conducts dropping the gate voltage of Q2, which in turn drops the current I1 the amount needed to allow the Vbe of Q1 to get back to equilibrium.
Here is a experiment.
Remove Q1 from the circuit.
Then measure the voltage drop across R3 with normal current flow (I1);
then increasse the current flow, and measure the drop across R3.
It Volt drop should increase across R3 linearly.
Now place a diode across R3 and apply enough current through the circuit to get a substantial reading across R3. Then increase current (I1) and continue to monitor the volt drop across R3.
It should remain fairly stable at the diodes forward voltage.
Now take out the diode and place Q1 back into the circuit leaving the collector unhooked for the moment. Again check volt drop across R3 with increasing current flow.
It should act like the diode :
Now connect the collector into the circuit and see how the measurements look then with increrasing current flow.
If you want to stop the current flow completely, with a threshold current increase, that will take more design work.

 

Ok ,
Now that you posted a circuit schematic,
From what you wrote and the schematic given,it looks like you using R3 as a current sense resistor, and the mosfet as a voltage controled current source, so if your current increases too much, you want the circuit to fold back and stop the current, this set up will not stop the current from flowing, as you noted in your description.
This circuit is acting as a feedback voltage regulator.
The reason the voltage is dropping at the base with increased current flow, is because Q1 is conducting at its Vbe value of 0.7v. , but what is happening as an increase of the voltage across R3 is sensed at the base, Q1 conducts dropping the gate voltage of Q2, which in turn drops the current I1 the amount needed to allow the Vbe of Q1 to get back to equilibrium.
Here is a experiment.
Remove Q1 from the circuit.
Then measure the voltage drop across R3 with normal current flow (I1);
then increasse the current flow, and measure the drop across R3.
It Volt drop should increase across R3 linearly.
Now place a diode across R3 and apply enough current through the circuit to get a substantial reading across R3. Then increase current (I1) and continue to monitor the volt drop across R3.
It should remain fairly stable at the diodes forward voltage.
Now take out the diode and place Q1 back into the circuit leaving the collector unhooked for the moment. Again check volt drop across R3 with increasing current flow.
It should act like the diode :
Now connect the collector into the circuit and see how the measurements look then with increrasing current flow.
If you want to stop the current flow completely, with a threshold current increase, that will take more design work.

ok.
I'm looking to create a constant current source with the circuit. But I find as the R3 resistor decreases, current does not increase as I expect(I=.7/R3).(turn on Q1 voltage over sense resistor), I find Vbe to be measured near .2-.5 volts and I cant get the current higher, to about 2.5amps where I want it.

 

I didn't feel like breadboarding this so I did a simulation of a mosfet circuit ,

Now here is the results of several tests,

I used a 1K resistor for the drain and a 1k resistor for R3.
The current flow was around 4mA.

Then I put a diode across R3, the current increased through Drain resistor (call RD) up to 9mA.

The voltage across R3 was 0.7 volts.

Now I decreased R3 by about 1/2, and Volt across R3 dropped, a little, current through RD remained around 9mA.

I decreased R3 around 10% more and the volt drop across R3 dropped more, but current through RD remained at 9 mA.

During these times most of the current was flowing through the diode and less through R3, so if I calculated current through R3 as ( Vdrop/ R3) it will be much less, but maximum current is still flowing through the parallel combination of diode and R3.

When R3 is at a lower value the voltage is going to show lower as it gets closer to ground, however the current remains the same throughout the circuit, because the diode acts as a volt ref. between gate and source, keeping the current fairly stable through the mosfet.
So even though your calculating a smaller current through R3, the remaining of the current is flowing through the base emitter diode as well.

The total current is the parralel of R3 and a diode.

The base emitter junction of Q1, is probably causing a lot of this phenomena, as well.

If Q1 Vbe was removed completely, than current increase would happen as R3 became lower in value.

 

This is a proven circuit topology. There is either a faulty power supply, a wiring error, a bad part, oscillation, or the MOSFET is not capable of Ids>2A when Vgs<<Vcc.

EDIT: Stupid question: Do you have your ammeter in the drain circuit, where it belongs?

 

This is a proven circuit topology. There is either a faulty power supply, a wiring error, a bad part, oscillation, or the MOSFET is not capable of Ids>2A when Vgs<<Vcc.

EDIT: Stupid question: Do you have your ammeter in the drain circuit, where it belongs?

I know, this is why it boggles me I cant get it to work.
I have my ammeter on drain. (i've made that mistake in the past, I've tried to double check all connections, looking for the silly mistake)

 

I replaced the mosfet and transistor, The circuit now appears to work when the ammeter is in place of the LED.(i didnt measure Vbe but the current went to 2.6 amps, (where I initially calc'ed I wanted it))

However when I solder in the LED the current only goes to 1.5 amps..
So the circuit appears to work without the LED, but with it -nope.

What could I try to fix this ?

 

Maybe I missed it if this was given earlier in the thread, but what voltage is each LED rated at, and how many of them do you have in the circuit?

You may be running out of voltage to drive the LEDs. Are you?

 

The OP has claimed one LED. There is a reason I asked for pictures of the total circuit, there are way too many assumptions floating around. This is a case where two pictures (top and bottom, LEDs included) are worth a thousand words.

 

The OP has claimed one LED. There is a reason I asked for pictures of the total circuit, there are way too many assumptions floating around. This is a case where two pictures (top and bottom, LEDs included) are worth a thousand words.

An LED part number would help a lot. T6 is not a part number, AFAIK.

 

Led :
http://www.cree.com/products/pdf/xlampxm-l.pdf

Sorry I will try to get a picture soon.
(a picture would show I have a MOSFET and the led on a heatsink. I have the r3 resistor set by resistors in paralell going thru 2 switches. Switches switch on power as well DPST. Npn transistor is soldered on right on the MOSFET pins)

I tried adding different size capacitors across the led, but it did nothing.