Widlar Current Mirror: How Does Q2 Connect to the Mirror and Share the Base Node?

Hello!
 

My question/s is/are regarding the attached schematic.


At the bottom of the schematic you'll notice that there is a current mirror attached to each diff bjt pair.

For the purposes of this question I circled one of the BJT's as red and the other as blue.
Red BJT = Q1
Blue BJT = Q2

The topology being utilised here is the Widlar topology. What I don't understand is how Q2 is connected to the current mirror.

Is the connection from Q2 going through Q1 or is that somehow also connected to Q1?

A bjt has 3 pins so I'm assuming it's going through Q1 and connecting to the node at the base of Q1.
 

Generally speaking, I don't understand how you can have more than one current source by just connecting BJT's to the base node of Q1. Why don't you need to have an independent topology for each respective current source?

Wouldn't there have to be a limit? If I keep adding BJT's to the base node of Q1. For arguments sake, let's say I add 1000 BJT's to the base node of Q1. Therefore, will I have 1000 different widlar current sources that I can design for? 

My key question is: "What are the limitations for creating 'cascaded' widlar current sources?" Maybe it's the Vcc powering the current source - too much loading from one widlar source?

Thanks in advance!

Kind Regards; a confused engineering student.
 

Hi W F,

From my point of view, the limitation depends on the parameters you'll have in which you need to consider the purpose of adding BJTs into the circuit that you are creating. Vcc can be considered as one, other than that is the system if its suits to the design system you are creating. These are just few of those practical limitations.

Hope it helps.

Best,
Neil

I suggest you download 'designing analog chips' by Hans Camenzind,
designinganalogchips.pdf
and read the section on current mirrors,
cheers,
Richard