• State Resolved
  • Replies 2 replies
  • Subscribers 33 subscribers
  • Views 50 views
  • Users 0 members are here
Support feedback
Options
Options
Related

TS3USB221A: Help with TS3USB221A used as dual-port USB MUX / “small loop” repeater — devices not enumerating

yunfeng wang
Prodigy 50 points
Part Number: TS3USB221A

Tool/software:

USB MUX.pdf

Hi TI team,

I’m using two TS3USB221A chips to build a USB MUX/“small loop” circuit.

Goal / intended behavior

  • When MUX_CTRL_USB = High: both USB ports (PC_USB and DEVICE_USB) should connect to the CPU (RK3588).

  • When MUX_CTRL_USB = Low: the two external USB ports should be shorted together through the two TS3USB221A devices, acting like a simple USB repeater/bridge (PC side Left right arrow DEVICE side), bypassing the CPU.

Actual result

  • When MUX_CTRL_USB = High (ports → CPU): both ports work normally.

  • When MUX_CTRL_USB = Low (repeater mode): the PC connected to PC_USB cannot enumerate the mouse/keyboard connected to DEVICE_USB. The PC shows no new device.

Design summary

  • 2 × TS3USB221A (high-speed USB 2.0 analog switch), each handling D+ / D−.

  • CPU side: RK3588 USB OTG (high-speed capable).

  • External ports:

    • PC_USB → connects to a host PC.

    • DEVICE_USB → connects to a USB device (mouse/keyboard).

  • VBUS on each port is routed (5 V present from the PC on PC_USB; 5 V supplied to DEVICE_USB when needed).

  • ESD protection on the connectors; 22 Ω series resistors near the controller; controlled impedance routing for D+/D−.

Symptoms / observations

  • With fast switching to repeater mode, the host PC does not detect any low-speed/full-speed device.

  • If I connect the device directly to the PC (no repeater), it enumerates normally.

Questions

  1. Is it valid to “short” two USB connectors through TS3USB221A (i.e., use the switches to directly connect D+ and D− of a host port to a device port), or does USB require an active repeater/hub/PHY for this use case?

  2. For USB role detection (DP/DM pull-ups/pull-downs & VBUS presence), do I need extra components or timing control when switching the path with TS3USB221A?

    • Do I need to gate or delay VBUS or the DP 1.5 kΩ pull-up so that the host can re-detect attach?

  3. Are there recommended application notes for using TS3USB221A in host Left right arrow device bridging scenarios (not only host/device Left right arrow CPU MUX)?

  4. Do I need dead time between disconnecting from CPU and making the PCLeft right arrowDevice path, to avoid partial connects?

  5. Any speed limitations (LS/FS/HS) or signal integrity constraints when cascading two TS3USB221A devices back-to-back like this?

  6. Would TI recommend a different device (e.g., USB hub, redriver, or dedicated OTG switch) to implement a reliable “PC Left right arrow USB device” pass-through?

Attachments (to be added)

  • Schematic PDF: TS3USB221A_MUX_Snippet.pdf

Thanks a lot for any guidance on whether my approach is valid and what changes are needed (pull-ups/pull-downs, VBUS control, timing, or a different part).

Best regards,

Wang

  • 0
    TI__Genius 17595 points

    Hello Wang.

    1. It is valid to short the USB connectors, but you are basically doing is cascading the two muxes (as you mentioned in question 5). This will negatively affect signal integrity.
    We have an application note on the topic for more information.  

    2. No need for additional components on the logic lines (SEL or /OE) as long as the voltage level and timing specifications can be met. 

    3. Any additional information on the TS3USB221A, other what is in the datasheet can be found under the technical documentation section in the product page.

    4. I am not sure how fast you are switching between the channels. You can try that and see if it resolves the issue. You should be meeting the timing specifications below.

    5. Cascading multiplexers doesn't really affect switching speeds. It mostly degrades bandwidth and increases the Ron. In other words it degrades signal integrity. 

    6. From the passive mux perspective, you can probe the USB 2 signal on the Machine-USB1 and Device-USB1 and check how much it was attenuated. 

    Probe points below in red.

    If you can provide that, we can continue the discussion according to your findings. 

    I am also looping in the active muxes team that might be able to recommend a suitable redriver or hub.  

    Thanks,

    Nir 
     

  • +1 in reply to Nir Gilgur
    TI__Genius 17595 points

    Hello Wang,

    I have discussed this with the redriver team and we concluded that using the passive mux solution is fine. 

    Thanks,

    Nir