# How to create a complex transfer function amplifier bjt based

I need to create an amplifier as shown in the first photo bellow, where we have a 20dB/dec rise and at the end of transfer function -40db/dec decline. I see we have a zero and then a double pole. A basic bjt amplifier was created with 2n2222 in PSPICE shown bellow,Its transfer function in the last photo is far from the transfer function i need. I can add another stages and mosfets or feedback to my bjt frst stage. could you recommend a structure which structure has a similar transfer pole zero transfer function structure as shown in the first photo bellow?

Thanks.

• You have a very wide bandwidth and you will need more stages. $\sqrt{170\:\text{Hz}\cdot 170\:\text{MHz}}=170\:\text{kHz}$. That's the center frequency. The fractional bandwidth is $\frac{f_{_\text{H}}\:-\: f_{_\text{L}}}{f_{_\text{0}}}=\frac{170\:\text{MHz}\:-\: 170\:\text{Hz}}{170\:\text{kHz}}\approx 1000$. So definitely a low-pass and a high-pass. – jonk Jan 23 at 8:56
• Hello Jonk,How can i see what is the bandwidth improvement regarding adding a stage? What about the double pole in the end ,How can i achieve it? Is there a structure in LTSPICE you can recommend which i could start build and analise to get such response? Thanks. – ron398 Jan 23 at 8:59
• Given the flatness and the incredible bandwidth involved I think this is a Butterworth and also I do NOT know which BJTs to attempt here. This is crazy on the high-end. Are you absolutely sure about that upper limit? – jonk Jan 23 at 9:28
• Hello jonk,my base stage is 2n2222,suppose i want to add another stage,even mosfet based,how can i make the double polein the end?(assuming we cannot reach such high bandwidth) – ron398 Jan 23 at 10:34
• It's a first order on the low end and a 2nd order on the high end. But the upper frequency is beyond my knowledge. I've never attempted anything that high with discrete construction. But I'm just a hobbyist. Maybe some expert here can tell you how. Your fractional bw is insane. Never encountered a problem that wide. You'll need to find more reasonable numbers if I'm to be of any good. I did write something here that may help a little, though. – jonk Jan 23 at 11:12