Library components that connect, run, and give the wrong answer

Four standard-library behaviours found while building the QX-250 power, rotation and redundancy models on Simcenter Amesim 2511 (EXERCISED 2026-09-24). Each passes connection validation and runs without an error; the only symptom is a wrong result. Check for them in the first run of any model that uses these components.

MECRL0 ignores windage unless friction is switched on

MECRL0 (icon rotaryload2ports) has a friction switch where 1 means no and 2 means yes, and the default is 1. With windage coefficients entered and the switch left at 1, the load behaves as a pure inertia.

Symptom: a DC motor driving it settles at exactly its no-load speed, V divided by the back-EMF constant. Set the switch to 2.

The eb diode's port 0 is the cathode

For EBDIODE01V01, port 0 is the cathode. Connecting a battery to port 0 reverse-biases the diode, and the load then draws current through its off resistance of 1 MOhm. The bus voltage goes to around minus 8 MV.

Symptom: an absurd negative bus voltage. Check its sign after the first run.

EB3N03 cannot join two diodes and a current load; EB3N06 can

EB3N03 has mixed causality: ports 0 and 1 give voltage and port 2 gives current. Joining two diodes and a current-source load with it fails at connection time with "ports are not compatible". EB3N06 gives voltage on all three ports and connects them.

The subtraction block computes port 2 minus port 0

JUN3M (sig icon subtraction) outputs input3 minus input1, that is port 2 minus port 0. Wired the other way round, a complementary filter in the QX-250 attitude model diverged to 96 rad (EXERCISED 2026-09-24).

The signal switch passes port 3 below the threshold

SWITCH01 (sig icon signal_switch) passes the value at port 3 while the command is below the threshold, and port 2 when it is above. Wiring it the other way round is easy and nothing flags it.