Fault Circuit
Background Information
***I have permission to discuss this project ***
Nuvation has:
1. A battery management system for stacks up to ~1500 V DC.
2. A cell interface per battery stack. The Cell Interfaces bolt to the cells and handle cell-level voltage/temperature measurement and balancing.
3. A cell interface monitors 24 series cells (one battery stack)
Problem Definition
Need to develop a companion tool for customers that do the wiring to detect wiring errors that could damage a G5 Cell Interface before you plug the real cell interface in.Goals:
1. Design a circuit which plugs into the cell interfaces’ connector which detects any miswiring fault
2. All components must be simple and inexpensive.
3. Limit the number of components to a minimum
Technical Approach and Design
Step 1: Identify the constraints and fault conditions
Each cell in the battery stack has a nominal voltage of 4.3V.
1. Normal Condition
- Any branch strictly below 8.6V (7V for headroom) is considered to be the threshold for normal condition and there should exist a clear indicator that there is no fault.
2. Overvoltage Condition
Any voltage detected above 7V to 103.2 V should clearly signal a fault
3. Undervoltage Condition
Any voltage detected below 0V to -103.2 V should clearly signal a fault
Step 2: Circuit Design
I went through many iterations in my design but unfortunately forget all the steps that I took to come to the completed circuit.
I took on an iterative approach. I tackled one fault condition at a time, and if this fix caused other conditions to fail I would identify why and fix it again
Overvoltage Detection Implementation
How does it work?
1. 0V -> ~7V the mosfet VGS < Vth so the LED is OFF.
2. 7V the mosfet begins to turn on and conduct current.
3. At high voltages 20V – 100V the current would be too high for both the LED and the MOSFET.
- A 4.7V zener diode clamps VGS to be a maximum of 4.7V to meet mosfet specifications.
- A 4.7V zener clamps the voltage drop across the LED to ensure current does not exceed ~6mA
- The resistor ensures that there is no greater than 15mA through the MOSFET (Ids).
4. A diode is added to prevent the LED being turned on in the undervoltage condition.
How did I define the component values/parts?
1. (Branch 2) Need to select a MOSFET with a high VDS rating. I selected 100V NCE0103M NMOS.
2. (Branch 1) Need to define the turn on point of MOSFET which conducts during the first overvoltage condition but off during normal condition.
The MOSFET selected has Vth of 1.5V – 2V. Therefore, it will start to conduct at the overvoltage condition.
But how much current will it conduct at this state needs to be determined. After lab testing it was conducting around 1mA which is sufficient to see the LED ON.
3. (Branch 1) Added a Zener Diode of 4.7V to clamp VGS at high voltages so the first resistor drops the voltage
4. (Branch 2) Used R22, LED4, with a zener clamping at 4.7V for a point at which the voltage may be too high for the LED, causing too much current.
5. (Branch 2) A source resistance of 200 ohms to limit the maximum current to limit the power seen by the mosfet < 1.5W
8.6V
100V
How does it work?
- Top diode prevents current from entering the branch during normal and overvoltage conditions
- Any negative voltage goes through the last branch and is limited to a maximum of 6mA. The LED with turn ON to indicate an undervoltage condition
