BJT Aleph input

[Koy]

After finishing the Aleph I started experiments with it (my recommendation - Aleph is easy to understand, lot of fun). I replaced the MOSFETs in input differential pait to BJT. But now I am not able to reach zero ( or almost zero) DC offset when using current mirror. Offset ( about 0.6 to 2V) moves with bias current. Why? Has it something to do with lower input resistance of BJTs? Or different "opening" voltage of BJTs? Why is not the pair with current mirror able to set 0V offset? Any literature to learn about?
Thanks!

 

[Nelson Pass]

Try loading the BJT's with resistors instead of
current mirrors and see if the problem goes away.

Also, BJT's with high current gain would be the best
for low offset.

 

[Mark A. Gulbrandsen]

What is a BJT? A Boxy Unijunction Transistor? I missed this one along the technolocical way someplace.....
Mark

 

[hifiZen]

Bipolar Junction Transistor... aka your standard garden-variety transistor, available in either NPN or PNP flavours.

 

[Koy]

NP:With only resistor - yes. But I think mirror offers more linearity. thats why I asked. Maybe low gain of input BJTs?

 

[Nelson Pass]

It is my experience that current mirrors offer more gain,
not necessarily more linearity.

 

[eLarson]

It is my experience that current mirrors offer more gain,
not necessarily more linearity.

What about a current "sink" at the source (emitter) pins of a common-emitter or diff-pair? Does that help linearity (akin somehow to emitter degeneration with a resistor, but with a much higher dynamic impedance)?

Erik