# Cascaded Filter Output Not Correct

I am simulating a speaker system with tweeter, woofer, and midrange speakers. I know my values for the components are all correct, but the output plot is a little off. The passbands should be a little below 0dB, but are at -20dB. The phase angle plots are correct though. Below is a screenshot of my circuit and the graph. Please help as to why this isn't working. • What you are plotting is the current for each Resistor. And the current is divided between your 3 branches here. You should plot voltage instead. – vrleboss Apr 27 '17 at 20:11
• @HervéGrabas, the current isn't divided between the branches since the filter passbands don't overlap (much). But the resistor value is 8 ohms, meaning 0.125 A delivered when each filter is ideally passing the signal. 20*log10(0.125) is -18 dBA. – The Photon Apr 27 '17 at 20:32
• @ThePhoton agreed +1 – vrleboss Apr 27 '17 at 22:48

The dB scale on the graph in LTSpice is relative to 1 V when measuring voltages or 1 A when measuring currents.

Your resistor values are 8 ohms, so 0.125 A AC is flowing through them in the pass-band of the associated filter

$$20\log_{10}{\frac{0.125\ {\rm A}}{1\ {\rm A}}} = -18.06\ {\rm dBA}$$

I'll bet if you put a marker on your graph, you'll find this is the value being reported.

how could I keep the resistors at 8 Ohms (it is required for the project), but get the passband to be closer to 0dB?

There are lots of ways to do this:

• Plot the voltage across R1 instead of the current through it (hat tip to Herve Grabas, in comments)
• Plot I(R1)*8 instead of I(R1)
• Set the magnitude of V1 to 8 V
• Plot I(R1)/I(V1)
• Replace V1 with a current source with magnitude 1 A.

The main issue is that LTSpice is sloppy. It's not really plotting a gain in dB. It's plotting a current or voltage in dBA or dBV.

• how could I keep the resistors at 8 Ohms (it is required for the project), but get the passband to be closer to 0dB? – python_ Apr 27 '17 at 21:27
• You're misunderstanding what this output means. In short, your passband is at 0dB, LTspice is just displaying it weirdly. Try measuring the voltage instead of the current! – Hearth Apr 28 '17 at 1:11
• How can I plot the voltage across R1 instead of the current through it? – python_ Apr 28 '17 at 19:35
• Click on the wire instead of the component. – The Photon Apr 28 '17 at 20:13
• @Felthry LTspice displays it accurately. What's weird is the user's interpretation. Or expectation. – a concerned citizen Apr 29 '17 at 6:07