BJT Miller Effect

[Marco Ferra]

Hi. I will be brief.

After testing the frequency response of the NPN BC337 in the lab I came
across the Miller effect. The gain at high frequencies dropped -3dB at
40.6 KHz. One of the questions of the report was to calculate the
internal parasitic capacitances C-pi and C-miu but it seems that I'm
only able, using Miller's theorem, to calculate the total CT parasitic
capacitance.

Is there a way, knowing the upper corner frequency, to obtain
mathematically and individullay C-pi and C-miu knowing the gain
K = (Vc / Vb) ? (btw this gain is negative because Vc(ac) < 0)

And is there a way, not to estimate the value of Rin, that I am making
Rin = Vb / Ib, but to know accuratly also the values of Rx and R-pi?

If I'm seeing this (what I think I know) in the wrong way please do tell
me.

Sincere regards, Marco

 

[Robert Baer]

Hi. I will be brief.

After testing the frequency response of the NPN BC337 in the lab I came
across the Miller effect. The gain at high frequencies dropped -3dB at
40.6 KHz. One of the questions of the report was to calculate the
internal parasitic capacitances C-pi and C-miu but it seems that I'm
only able, using Miller's theorem, to calculate the total CT parasitic
capacitance.

Is there a way, knowing the upper corner frequency, to obtain
mathematically and individullay C-pi and C-miu knowing the gain
K = (Vc / Vb) ? (btw this gain is negative because Vc(ac) < 0)

And is there a way, not to estimate the value of Rin, that I am making
Rin = Vb / Ib, but to know accuratly also the values of Rx and R-pi?

If I'm seeing this (what I think I know) in the wrong way please do tell
me.

Sincere regards, Marco

Rin, if i remember correctly, is beta*re+rbb' (that is little Remitter
and base spreading resistance); and re is a linear function of current
(it has been 30 years but i think 27 ohms at 1mA - correct me if i am
wrong).

 

[Marco Ferra]

Rin, if i remember correctly, is beta*re+rbb' (that is little Remitter
and base spreading resistance); and re is a linear function of current
(it has been 30 years but i think 27 ohms at 1mA - correct me if i am
wrong).

Yes. You are right. In fact re = VT / IE.

(being VT bjt's thermal voltage = 25 mV so 25m / 1m = 25 ohms).

The report is ambiguous about this "Rin" but I guess if i make my point
and prove it mathematically all will be fine.

Thank you for the reply
Marco