I noticed if I change AC amplitude for AC analysis the gain changes.
What AC amplitude is the best to use to test for AC analysis to determine gain in general, or is it an end application parameter.
AC1
AC10
AC1m
\$\begingroup\$
\$\endgroup\$
2
-
3\$\begingroup\$ Freddy, \$M_1\$ won't really drive \$R_\text{L}\$. As a source-follower it can source current into the load, but it can't sink it. The only thing doing that is \$R_{12}\$ and that is much too large in value. The capacitor bypasses at the other stages mean excessive gain and distortion. The BJT biasing pairs are set too high in value. Other problems, as well. The basic cookie-cutter approach makes it appear better designed than it is. Where are you at in your education? What are the specs for this, other than source and load impedances? \$\endgroup\$– periblepsisCommented Apr 28, 2023 at 20:54
-
1\$\begingroup\$ Your MOSFET is a defined as a depletion-mode device instead of enhancement-mode. I've never seen that used in such a configuration. Is that intentional? \$\endgroup\$– Ste KulovCommented Apr 29, 2023 at 2:16
Add a comment
|
3 Answers
\$\begingroup\$
\$\endgroup\$
1
You can use 1V in all situations.
You'll get the same results as with different voltages except the gain will be correct (at least the gain will be correct for small inputs, not necessarily for 1V).
SPICE finds the operating point and then linearizes about that operating point so the amplitude is not used in the analysis except to scale the output (so if it saturates, distorts, clips etc. that is not modeled). For that reason, it's best to do a transient analysis before believing the AC analysis.
answered Apr 28, 2023 at 20:19
-
\$\begingroup\$ Got it, I'll try as a comparison. Thanks. \$\endgroup\$ Commented Apr 28, 2023 at 20:43
\$\begingroup\$
\$\endgroup\$
The gain is the same in all three sims.
The gain is V(output)/V(input)
Therefore, the correct value is the one from the AC 1 simulation.
\$\begingroup\$
\$\endgroup\$
5
I agree that your Mosfet follower is severely overloaded. Your input signal level is way too high. The gain and distortion are also very high even when the input level is turned down.
Here is only your 1st stage:
-
\$\begingroup\$ Note that this is not important in an AC analysis because this is a small signal analysis in which the circuit is linearized around the operating point, and the AC amplitude doesn’t matter. The problem in the OP is just that they are determining
V(output), and not the gain. \$\endgroup\$ Commented Apr 29, 2023 at 14:28 -
\$\begingroup\$ I agree that the total gain is about only 13200 (91dB) because of the severe overloading of the Mosfet follower working as a rectifier. \$\endgroup\$ Commented Apr 30, 2023 at 16:15
-
\$\begingroup\$ 91 dB is not the gain, it's the output voltage (in dBV) when the input voltage is 10 V, under the assumption that the amplifier is linear, because that's how AC analysis works. Since 10 V corresponds to 20 dBV, the gain is actually around 90 dBV - 20 dBV = 70 dB, as can be seen from the OP in the plot with the analysis with AC 1. The AC analysis does not consider any saturation. You can input 10000 V and it'd give the same gain. \$\endgroup\$ Commented Apr 30, 2023 at 16:23
-
\$\begingroup\$ But the amplifier that is severely clipping and its output Mosfet is severely overloaded are not linear. Then why simulate it? It should be simulated the way it will be used. \$\endgroup\$ Commented May 1, 2023 at 17:22
-
\$\begingroup\$ The OP is doing a specific type of simulation, which is not the one you're using in your answer. Every type of simulation gives certain pieces of information, and the important point is learning how to interpret them. So, again, what the OP is observing has nothing to do with saturation because the AC analysis is a linear analysis and you can set the input amplitude as large as you wish. It's useless, but it gives the gain anyway. \$\endgroup\$ Commented May 1, 2023 at 17:27