# LTSpice Question

Discussion in 'Electronic Design' started by Gary Richardson, Jan 15, 2004.

1. ### Gary RichardsonGuest

I've created a LTSpice simulation of the bandpass filter of Figure 5.19 on
page 278 of A of E. My circuit is shown below; the center frequency is
60kHz.

In the LTSpice simulation the gain at the center frequency is +9 dB and the
Q is about 60; the expected values are 0 dB and 20. However, the simulation
using Ansoft Serenade (Student Version) does produce the expected values.

I'm an LTSpice newbie so I probably have something screwed up in my
simulation but I have no idea what it might be. Any suggestions or comments
appreciated.

Gary Richardson

----------------------------------------------------------
State-variable bandpass filter. A of E, page 278
R1
+--------------/\/\-----------------+
| 100k |
| |
| Rq R C |
+-/\/\-+ +--/\/\-+ +--||--|
| 5k | | 10.0k | |1.0nF |
| | | | | |
Rg | |\ | R | |\ | Rf | |\ |
In -/\/\-+-|-\ +-/\/\-+--|-\ +-/\/\-+-|-\ | Out
100k | >-+10.0k | | >-+ 2.65k | >-+---O
+-|+/ |+-|+/ +-|+/ |
| |/Q1 || |/Q2 | |/ |
V |V V Q3 |
| C |
| +--||-+ |
f = 1/2 Pi Rf C \ |1.0nF| |
R/10.0k | | Rf |
Q = R1 / Rq \ | /| | |
| | / |-+-/\/\---+
G = R1 / Rg +------+< | 2.65k
\+|-+
Q1, Q2 - LT1498 \| |
Q4 V

2. ### Mike EngelhardtGuest

Gary,
The unexpected peaking is due the the 1498's not having
infinite GBW. There's a opamp guideline of 3*Q minimum
gain at resonant freq. for statevariables in D.L.'s _Active-
Filter CookBook_, but this version of the circuit seems to
need more a times times more than that to accurately
duplicate the desired gain and Q. Since 3*Q is 35dB but
the 1498 model has 44dB at 60KHz.
That would have been with a different opamp or model.

--Mike

3. ### Gary RichardsonGuest

[snip]
Thanks for the info. I ordered a copy of the "Cookbook" today.