Joule Thief

[stprue]

Fantastic vid slayer!  you should make a switching circuit to see how long it will run for and also to see if the charge going into the batteries is static or not!

[mscoffman]

Can any of you folks give me links or tips on how to use mosfets.  I'm having trouble figuring out how to use one like a voltage controlled resistor.

I would use a triac or thyristor but they have some junction issues and tend to go into runaway too easily.

---
Triac and thyristors and SCR's are good for higher voltage and high current
loads but their input energy needs to be AC because their *load* voltage
has to go to zero volts for them to turn off. In other words they are
triggered on and wait to be cut off and do not gate on and off. They are an
excellent choice for high power and low frequency AC where applicable.
They don't particularly run away except, where they are often used,
they are run at extremely high power levels.
---
Bipolar transistors are current gain devices. Their transfer equation is:

Ic=B*Ib

where Ic=collector current, B=transistor Beta spec, Ib=base current
Beta (gain) is between 20 for power transistors and 100 -> 200 for
2N2222 sized transistors. Current through the transistor happens
when the base is positive with respect to the emitter for NPN's
and current flows out of the base from the emitter for PNP's
---
Mosfet transistors are voltage gain devices. Their transfer equation is:

Es = A*Eg

Where Es=Source Voltage; A = mosfet gain; Eg=Gate Voltage
The suprising thing is A gain may easily be 10^6, a million or
more and very little or no gate current is required to turn them on.
So they are excellent for use as digital switches. Which is how
you usually see them used. They have packages available that
parallel many silicon devices that can bring their on-resistance
down to .01 ohms!
---
One problem is when they are used to switch voltages at five
volts or below the source to drain voltage the gate wants
to see is a volt or two above the source voltage for the
transistor to fully saturate on! In this case it is recommended
to use a mosfet driver IC that can generate the required gate
voltage to fully saturate the mosfet transistor!
---
So how do you use them linearly? You have to use "negative
feedback", where a negative voltage is summed with the
positive gate voltage to actively bias them into the middle
of their range. Summed means using an inverting amplifier
and summing resistors.

The other thing you can do is to use them as high frequency
switches and use PWM pulse width modulation or PPM pulse
proportional modulation to control voltage integrated on
a capacitor. This can form the basis of a switching amplifier
where the output mosfets switch on/off at frequencies much
above the frequencies being amplified. These switching type
amplifiers can be made highly energy efficient in comparison
to the bias method that "wastes" a lot of power in the transistor.

:S:MarkSCoffman

[innovation_station]

: )

LOOK AT MY FAMILY AND THE SONGS THEY SING FROM THERE HEARTS! 

: )

DO YOU KNOW HOW GOOD THIS MAKES ME FEEL!

: )

WILLIAM 

am i your anti christ ?    answer no ....   you are mine !   i am you !

http://www.youtube.com/watch?v=HMxqiNII6KU&feature=fvw

Bob Taylor - Deja-Vu ( Radio Version By Play & Win )

[stprue]

Here is my recent vid!  I have sense improved the circuit so that I don't need to touch it and in fact I have added a couple of LEDs on an additional av plug and I can reduce the ma draw by adding LED's! 

So more LED's= less MA draw

http://www.youtube.com/watch?v=jTI-bfXFGzk

[jadaro2600]

---
:S:MarkSCoffman

Thankyou sir, excellent advise.

I also found a few devices called VCR, which do exactly that.  The voltage controlled resistor actually has a pinout like a transistor; I think I may investigate these some more.

My issue is, having very low voltages to deal with.  Most of the transistors don't fully cut on with just 1.2 volts, there's still a good bit of resistance in the device.  I also noticed, when testing my transistors, that the lower the base voltage ( or resistance to base ) the more resistance across the C-E junction ...then this decreased as base resistance increased, up unto a point when it started increasing again.  The strange thing was that it dipped.

I'll have to post a table or something later.  I'll also have to test multiple transistors to see if this is phenomenal or just a fried transistor.