Electrical Training Board

In industrial control, adding resistance into circuits would do the job.

When <...current below normal...> refers to 4-20mA <...analog...> control circuits, a handful of resistors in the 1kΩ 0.5W region wired in series in the circuit can normally clip the current to below 20mA, and drive the low end below 4mA. However, the whole idea of a live zero at 4mA is that a wire break will show up as a low current alarm.

The <...voltage below normal...> in <...digital...> input circuits is a different matter. The <...sensors...> ought to be wired failsafe, such that a closed contact is a normal signal and an open contact is an alarm/abnormal one. Disconnecting the circuit would then qualify.

However, in other circumstances such as <...electronic engine control...>, the above might not be applicable.

Can potentiometers be used in engine control systems to create these faults?

The first step is to get on the phone to the original equipment manufacturer(s) and ask there. If circuit diagrams are not already to hand, then a little gentle persuasion can be employed directly, rather than hanging about endlessly on CR4.

Make some calls.

You need some sophisticated test machines...as well as the training to use them....

..."Vehicles are packed with sensors that feed into the on-board diagnostic (OBD) system and ECU (see fig. 1). Those communications take place using serial data links. The Ethernet, Controller Area Network (CAN), Local Interconnect Network (LIN), Media Oriented Systems Transport (MOST), FlexRay, and Peripheral Sensor Interface 5 (PSI5) protocols are among those found in such applications, as well as others based on the foundation of the Manchester line-encoding scheme. [Also see Decoding Manchester-Based Serial-Data Protocols For Sensor-to-Controller Applications]

A system connecting myriad sensors to an ECU using the PSI5 protocol might resemble the block diagram shown in Figure 2. Sensors connect to the ECU with just two wires, using the same lines for power supply and data transmission.

The transceiver ASIC provides a pre-regulated voltage to the sensors and reads in the transmitted sensor data. The example above shows a point-to-point connection for sensor 1 and 2 and two different bus configurations for sensors 3 and 4 and 5 to 7, respectively.

To properly design and debug communication between ECUs and sensors, be sure that your test instruments are properly equipped to decode the serial data stream. One example is Teledyne LeCroy's High Definition Oscilloscopes (HDOs), which can be configured to decode many different serial protocols. In general, the LIN protocol does a good job of serving low data-rate applications and CAN handles medium-speed requirements. The MOST protocol is tailor-made for high-speed tasks while FlexRay serves safety-critical applications such as steer- and brake-by-wire.

One interesting capability offered by the combination of modern oscilloscopes and associated software is the ability to create a virtual analog waveform from digital serial data. For example, Teledyne LeCroy's CANbus TDM (M=Measure/Graph) and ProtoBus MAG (MAG = measure, analysis, graph) packages provide a parameter that is essentially a digital-to-analog converter. This parameter extracts a serial data message and scales it to an analog value. If there are multiple analog values, as there would be in a serial-data sensor stream, the software presents the values as a waveform, in turn creating a virtual analog waveform from the digital serial data."...

https://www.fierceelectronics.com/components/test-considerations-for-automotive-sensors

https://www.electronicdesign.com/magazine?

I don't understand your project. If you have removed all the sensors, wiring harness, and Engine Control Unit (ECU) from a vehicle(?) the troubleshooting should be easy and obvious once a visual inspection is performed. The entire electrical system is missing!

If instead, you want us to explain all of the signals the vehicle sensors should simultaneously produce then you have a problem. That is an enormous request that many here (myself included) would tackle only if we were hired as a consultant. I would also insist on knowing the vehicle's make, model, and year vehicle to be simulated.

You also use the word "training" in an ambiguous fashion. Who or what are you training with this board; the sensor fabricator, the ECU, or the service technician?

Well, maybe you should start with adding a OBD port (On Board Dianostics) to your training board…

Since most faults are not sensor failures, I would employ typical failure modes, using mechanical means, a switch with a rotating brush on a rusty contact plate, two frayed wire ends tied with rubber coupling you could simulate road bumps with, blower for your MAP and air velocity sensors, coordinated with O2 and fuel demand signals, to simulate bad O2 sensor or MAP or air flow, little rotary hammer drill to excite your knock sensor, O2 and engine torque adjustments to simulate bad spark coil or plugged injector.

You may find that you will get a better or more useful educational experience by tricking out an actual vehicle, with bad sensors or less than ideal performance parameters, worn rings, valves, bad EVAP purge valve, fuel level / pressure sensor, rather than the elementary exercise of learning to operate a digital multimeter.