Two stage amplifier (opamp adder + common emitter), anomalies PSpice: bandwidth increase

Thread Starter

AndreaM

Joined Apr 19, 2017
11
I am designing a two stage amplifier, the first stage is a opamp adder and the second stage is a common emitter (see figure).
I have designed the two separately, for the opamp the specifications are Rin=50 Ohm, two source, A=1.
For the common emitter the specifications are Rout=50 Ohm, A=20, Fl=200 kHz, Fh=1 MHz (at least).
Some parameters:
Vcc=+6 V, Vee=-6 V, Beta=170, Gm=0.4 S, Ic=10.04 mA, rbe=420 Ohm, ro (common emitter) = 7968 Ohm.

I have simulate the two separately, Vs1=0.5 V(AC), Vs2=0.5 V(AC), Vs3=1 V(AC) and the results are (figure):

1) opamp
Fh = 1.52 MHz

2) common emitter
Fl = 150 kHz
Fh = 573 MHz
A = 14

Then I have simply connected the two (figure) and I have measured the overall results:

Fl = 127 kHz
Fh = 1.79 MHZ

those result are in contrast with the theoretical expectation, I expect that connecting two stage the bandwidth decrease but this does not happen on the simulation. The overall Fl is lower than the maximum of the two separately and the Fh is higher than the minimum of the two.

So where is the error? Does PSpice have some particular attribute to beware?
 

Attachments

Is there something about the graphs that looks out of place? I think you are getting results you do not expect because there is no long flat portion of the final output. That is, the HPF portion interferes with the LPF portion and vice versa. The interaction is skewing the calculated result.
 

Thread Starter

AndreaM

Joined Apr 19, 2017
11
Is there something about the graphs that looks out of place? I think you are getting results you do not expect because there is no long flat portion of the final output. That is, the HPF portion interferes with the LPF portion and vice versa. The interaction is skewing the calculated result.
What HPF and LPF stay for?

So you guys think that may be correct? My professor says it is not possible, we have did a theoretical analysis that I upload (if someone interested) and the conclusion is the overall hight cut-off frequency should be lower than the lowest of the single stages.
 

Attachments

LPF and HPF are just low pass filter and high pass filter.

If you have only LPF or HPF elements then your math works out fine. However, you have both in your circuit. To make matters worse for the math, you have placed the cutoff frequencies close to one decade apart. One filter has a rising slope in the same area that the other has a falling slope.

You do not reach the pass band, or peak, of one filter before you begin to influence it by the other filter. This changes the calculated 3 dB points calculated by the software.
 

Thread Starter

AndreaM

Joined Apr 19, 2017
11
Thanks, now I get it. Your argument is valid, I tried to make the capacitors 100 time bigger to have about 100 time smaller lower cut-off just to test this and like you said in this situation I found the expected result.
You earned a like!

Do you know how may I analyze the LPF + HPF situation with math? If you have some advice or if you can write it (least the important parts) would be really appreciate. I have the "Microelectronic Circuits 7ed Sedra and Smith" I have not seen this analysis on here but if I'm wrong and someone know the paragraph would be fine.

Furthermore I have another anomalies on this project about the lower cut-off frequency that is not like the paper project, I have posted it on "The Project Forum", link here. Can you please take a look?
 

Jony130

Joined Feb 17, 2009
5,211
I do not see any "anomalies"? As we can see on the blue plot the gain is lower and this is why Fl and Fh look slightly lower.
But If you will do the full analysis without any approximations, you will confirm PSpice results.
 
I think you will have to go into the frequency domain or s domain to find your answer. Find transfer functions for both filters and apply both at the same time. This is definitely not my strong suit here so maybe someone else can chime in on the details of the operation.

I will look at the other post when I get a chance.
 
Top