question about a PNP current mirror (BJT)

[rmanor]

current mirror

Hi,

When I look at the an NPN current mirror (implemented with BJTs), I understand why the diode connection (base and collector are short circuted) is required at the master transistor - we need to copy the voltage at the collector to the base in order to have the Vbe voltage in the active mode (usually > 0.7V).
But in the PNP current mirror, as far as I can see), even if I don't do the diode connection there is a sufficient Vbe voltage ( < -0.7 for PNP).
Where am I wrong? or is the connection there for a different reason?
thanks.

 

[FvM]

pnp bjt current mirror

Where do you get the base current for both transistors without a connection to the input? Actually, there could be |Vce| > |Vbe|, but 1. there must be a connection and 2. the input transistor collector current must be able to shut down the base current to allow the current mirror to operate. There is no difference between NPN and PNP current mirror at all, except for current and voltage polarity, I think.

 

[rmanor]

current pnp mirror

Where do you get the base current for both transistors without a connection to the input? Actually, there could be |Vce| > |Vbe|, but 1. there must be a connection and 2. the input transistor collector current must be able to shut down the base current to allow the current mirror to operate. There is no difference between NPN and PNP current at all, except for current and voltage polarity, I think.

wont I get a base current from the collector/emitter voltage?

 

[FvM]

No, I don't think so.