SCEA147 September   2024 TXB0104 , TXB0302 , TXB0304

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1Introduction
  5. 2TXB-Type Architecture Differences
    1. 2.1 TXB0104 and TXB0304 Comparison
    2. 2.2 Recommended Device Selection
  6. 3Common Design Challenges
    1. 3.1 Distributed Load (Transmission Line Effect)
      1. 3.1.1 Cable Length Impact - Bench Findings
    2. 3.2 Impedance Matching
      1. 3.2.1 Bench Setup and Results
    3. 3.3 Lumped Load
  7. 4Case Study
  8. 5Summary
  9. 6References

Cable Length Impact - Bench Findings

Shown are bench waveforms captured with TXB0304 connected to a 20+ in. and 4 in. cable length. Note the same input conditions were applied to both setups.

 TXB0304, 1.2V Input (Dark Blue) to 2.5V Output (Teal) With 20+in. CableFigure 3-3 TXB0304, 1.2V Input (Dark Blue) to 2.5V Output (Teal) With 20+in. Cable
 TXB0304, 1.2V (Dark Blue) to 2.5V Output (Teal) With 4 in. CableFigure 3-4 TXB0304, 1.2V (Dark Blue) to 2.5V Output (Teal) With 4 in. Cable

From the waveform captured in the 20+in. cable above, the input (Dark Blue) oscillates due to unintentional re-triggering of the one-shots at both the input and output sides. Recall that TXB architecture is similar to a buffer loop, with I/Os always being redriven to a known and similar state. When paired with long cabling however, the I/Os result in different logic states due to the timing delay caused by the 20+in. cable. The result is similar to the observed oscillations.

When the setup was swapped to a shorter cable (4in.) in Figure 3-4, the output signal was now more easily interpreted with no oscillations.