TINA/Spice: Apparent error in simulating a passive circuit.

Tool/software: TINA-TI or Spice Models


I am simulating a passive circuit for a high-voltage differential probe. The circuit consists of resistors, the distributed capacitance to the ground plane, the input capacitance of the load and a trimmer capacitor to tune out the distributed capacitance and the input capacitance to the load. The DC path from signal source to the load consists only of series resistors of value 125k and one load resistance of 5k . Thus, there is no gain and there are no shunting resistors of any sort. One capacitor is in parallel with each resistor, and there is a capacitor to ground at each connection between resistors.

When I run an AC analysis and select the "AC Transfer Characteristic", I select a frequency sweep from 1 Hz to 100 GHz.

To my surprise, when Tina-TI simulates the circuit, it reports a gain of -37.6 dB from 1 Hz to 1 Megahertz. However, the hand-calculated gain is actually -40.08dB because the four 125k resistors in series total of 500k, and they feed a load resistor of 5k so the result is a 5k/505k divider which has a DC gain of -40.08 dB.  The capacitors play no part because they have values in the low pF range and the frequency at the lower end of the sweep is much too low.

What is wrong here? I can provide you with my .tsc file if you tell me how to get it to you.