supply independent current source, why forcing the same voltage?

I failed to understand what is the reason we want V1 = V2. What is the reason or motivation?

I am studying from here min 12:49 and here page 26.



appreciate any help.

The strategy can be duplicated to several circuits. Say you have many bulbs and you want to vary their brightness simultaneously with a single adjustment. You only need to make one current mirror, then attach multiple branches to that one current mirror.

I think what you said is the explanation why we use current mirror, to copy corrent. this I get it.

my question is in the circuits I attached above, we want to create reference circuits actually, one which is independent of supply, it is said that V1 = V2. this is the part that I do not understand.

Right, my example does not necessarily need your entire circuit.
Let's see... I can imagine a need to measure voltage at a sensitive (or delicate, or high impedance, or non-solid) component, and send that voltage to equipment with no connection to the original component (in order not to tamper with the original source).

The instructor also mentioned band-gap circuits. These are a complicated network which appears built from several of your schematic. This forum has discussions about bandgap circuits or bandgap references.

Hi,

At a guess, if V1 = V2 then I1 = I2 then the current source is balanced and stable. Think of it as a balancing act, or a see-saw, or a rocking boat - up on one side is down on the other. I might be wrong but I play with current sources s frequentlyand the Wilson one isn't as good as might be wished for.

IMHO, assuming the mysterious bubbles are resistors or similar, that is not a truly supply-independent current source, at all, and nor is it temperature-independent, either. Greater Vin = greater I1 & I2 = higher V1 and V2.

If you want to see a super supply-independent currrent source, look for older versions of lm2903, lm339, etc. comparator datasheets for link to app note describing how those devices source is constructed, truly superb thing.

- - - Updated - - -

Hi,

If you are making a reference from a Wilson mirror, good to ponder PTAT and absence of CTAT...

My mobile signal is too weak to upoad the app note I mentioned above. Here's the link: snoa654A It'll possibly offer to download a pdf, it is from TI, safe...

Document is "AN-74 LM139/LM239/LM339 A Quad of Independently Functioning Comparators", aka snoa654A - May 2004 - Revised April 2013. Current source explanation is bottom of page 4 to midde of page 5.

You need to force the same voltage V1=V2 in order to termally compensate. Referring to pag. 26 of the lecture you indicate, you can see the two bottom bjt, diode connected. These have two different dimensions, let's say we have a factor "N". Due to this the cutrrent into the two bjt junction will be:

Iq1 = Is*exp(Vq1/Vt)
Iq2 = N*Is*exp(Vq2/Vt)

where Vt =K*T/q

solving for the voltages:

Vq1 = Vt*ln(Iq1/Is)
Vq2 = Vt*ln[Iq2/(N*Is)]

but we know Iq1 = Iq2 = I, so writing the equation referred to Vo1 and Vo2 (slide at pag. 26)

Vo1 = Vt*ln(I/Is)
Vo2 = Vt*ln[I/(N*Is)] + R*I

But if we force Vo1=Vo2 then:

Vt*ln(I/Is) - Vt*ln[I/(N*Is)] = R*I

LSH the voltage difference between the two diodes, RSH the voltage across the resistor (Vr).

by simple manipulation:

Vt*ln(N) = Vr

this means that the effects of the temperature (remember Vt=KT/q) are linearly reflected on the resistor acting the temperature compensation