• State
  • Locked Locked
  • Replies 0 replies
  • Subscribers 46 subscribers
  • Views 111 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TIDA-00929: Technical Inquiry: Arc Fault Detection System Design using 600/5A CT and TIDA-00929 Architecture

Halil Tuncer
Prodigy 10 points

Part Number: TIDA-00929
Other Parts Discussed in Thread: OPA835

Hello TI E2E Community,

I am currently developing an Arc Fault Detection (AFD) system. I am using a Current Transformer (CT) with a 600/5A ratio and a bandwidth of 1 Hz to 200 kHz. Due to cost constraints, Rogowski coils are not an option for this stage. I have been studying the TIDA-00929 reference design and would like to seek expert advice on several critical design aspects:

1. Frequency Band Separation and Dual-Purpose Path Design

The system must perform both RMS current measurement at 50/60 Hz and high-frequency (HF) component analysis for arc signature detection. I am planning to implement a dual-path architecture (LF + HF) similar to TIDA-00929:

  • LF Path (< 100 kHz): Can a single Sallen-Key LPF + ADC path reliably handle both fundamental RMS measurement and frequency-domain arc analysis? Or is it recommended to have a dedicated, independent path for high-precision metrology?

  • Crossover Frequency: In TIDA-00929, the LPF cutoff is 100 kHz, and the HPF starts around 100 kHz. Considering my 600/5A CT’s transfer characteristics, should these boundaries be adjusted, or is this 100 kHz split a standard "sweet spot" for AC line monitoring?

  • Isolation/Decoupling: How can I ensure that the large 50/60 Hz component and potential DC offsets in the LF path do not saturate or interfere with the high-gain HF path? What are the best practices for decoupling these two paths at the AFE input?

2. Sufficiency of the 200 kHz Bandwidth

TIDA-00929 specifies a maximum of 100 kHz for frequency-domain analysis while utilizing a comparator for time-domain signals up to 10 MHz.

  • Detection Reliability: Is a 200 kHz CT bandwidth sufficient for reliable arc fault detection in residential/industrial AC grids?

  • Nuisance Tripping: To distinguish "operational arcs" (e.g., motor start-ups, brush noise from drills) from "fault arcs" (series/parallel), are MHz-level components essential? Or is the signature below 200 kHz (e.g., "shouldering" effect in time-domain and broadband noise in frequency-domain) typically sufficient for robust discrimination?

3. AFE Modifications (OPA835 and Burden Resistor Selection)

The reference design is optimized for a ±155 mV peak input. Given my CT has a 5A secondary output:

  • Burden Resistor: What is the recommended value/range for the burden resistor to maintain a linear operating region while ensuring a high enough Signal-to-Noise Ratio (SNR) for low-level HF noise?

  • Gain Structure: How should the first-stage differential gain of the OPA835 be configured? Specifically, what gain is recommended to prevent HF noise components from being lost in the floor while staying within the ADC's dynamic range?

  • Saturation Management: Since I am performing current measurement and arc detection simultaneously, how should the gain be structured to prevent ADC saturation in the LF path during high-current transients without sacrificing HF sensitivity?

I would appreciate any insights, simulation files, or modified schematics that could assist in adapting the TIDA-00929 concept to a high-current 5A secondary CT application.

Best regards,

Halil