SLOA358 July 2025 DRV2605L
After characterizing the Audio-to-Haptic mode alone, the behavior when switching between Audio-to-Haptic (A-V) mode and manual drive modes was examined. In many applications, the system can normally run in Audio-to-Haptic mode for ambient feedback, but then temporarily switch to a direct haptic mode (like real-time playback or playing a library effect) when a specific event occurs that has a dedicated rumble pattern. The DRV2605L supports on-the-fly mode switching through the Mode register, but this is important to manage the transition to avoid any undesirable artifacts (like a momentary jolt or a pause in vibration).
Several scenarios were tested: switching from A-V mode to Real-Time Playback (RTP) mode, and vice versa, under conditions with and without ongoing audio input. The key observations were:
Understanding the Modes: In Audio-to-Haptic (A-V) mode, the DRV2605L autonomously drives the LRA from the audio input – no explicit I²C commands are needed to generate vibrations, as long as audio is present. In Real-Time Playback (RTP) mode, the DRV2605L instead takes direct commands: the host writes an 8-bit value to the RTP input register (0x02) at any time to specify the desired drive strength (effectively commanding the PWM output on the LRA). RTP mode is useful for generating arbitrary waveforms or playing back effects under host control. There is also a Waveform Sequencer/Library mode where the host can trigger predefined effect sequences (by writing to registers 0x04–0x0B for sequence and setting mode 0x00), but that is conceptually similar to RTP for the purpose of switching (the driver isn’t using the audio input in that case).
For example, Figure 5-1 captures a scenario where the device was initially in RTP mode (with no vibration output commanded), and we switched to Audio-to-Haptic mode while a 100Hz audio tone at 50% volume was already playing into the DRV2605L. The top half of the scope trace shows the blue audio waveform continuing through the switch. The bottom (red) trace is the LRA output. You can see that once A-V mode is enabled (around the switch moment), the output starts oscillating and within 1 cycle reaches the expected amplitude for that audio input. This confirms the device smoothly began audio-driven operation (with just a minor initial dip due to the algorithm’s settling).
Scope capture of switching from RTP mode to Audio-to-Haptic mode while a 100Hz audio tone is playing. Time 0 on the horizontal axis is the moment of mode switch. The blue trace (CH2) is the audio input (14.2mV RMS, about 50% volume) which is continuous before and after the switch. The magenta/yellow traces (LRA differential output) were flat during RTP mode (no drive commanded), then begin oscillating once Audio-to-Haptic mode engages. A brief amplitude ramp-up is observed immediately after t=0, as the DRV2605L’s audio algorithm takes over. By about 40–50ms after the switch, the vibration output has reached steady-state, matching the audio’s amplitude. The transition is smooth, with no large spikes or discontinuities.
Figure 5-1 Switching From RTP Mode to Audio-to-Haptic
Mode While a 100Hz Audio Tone is Playing
Figure 5-2 Switching From Audio-to-Haptic Mode to RTP
Mode, Using the Technique of Zeroing the Drive at Switch-overIn summary, to achieve glitch-free transitions between automatic audio haptics and manual control, the host needs to manage the RTP register value at switch time (either matching the current level or setting this to zero for an intentional pause) and be mindful of the short mode-switch delay. With these practices, the DRV2605L can seamlessly hand over control without the user feeling any discontinuity in vibration.