Matching transistors - Measuring hfe

[mchambin]

I just wanted to point out that there are other factors which may force the LTP from a balanced condition even though everything is matched in that location.
-Chris

Bumping an 11 years old topic....
Because of NFB; The unbalance of the LTP should be very small in the case of a large loop gain at DC.

 

[anatech]

Hi mchambin,
Still, any unbalance condition will increase distortion. Once you're past a certain point the distortion rises rapidly with lack of balance increases.

The best performance occurs when the rest of the design is balanced and tends to run the DC offset to zero. Then highly matched diff pairs can really make a difference. The diff pair will naturally try to correct for any non-linearity including a DC offset. Small DC offsets do not adversely affect the operation of an amplifier, so if there is a trade-off to be made, design so that the diff pair remains balanced in normal operation.

-Chris

 

[mchambin]

BTW, which is more important to balance a LTP: Hfe or Vbe ?
You can hardly have both. What is the priority ?

 

[anatech]

Hfe. Vbe will not be very different between Hfe matched transistors and you may well be able to achieve both. My own experiments have shown that a matched Hfe to give the lowest distortion, and technical books on transistors support that for the most part.

Matched Vbe only matters in temperature sensors where you are measuring differences in Vbe between two or more points. Other than cases like that, I can't even understand why this is a question.

-Chris