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.

MUX36S16: USB mux/switching?

Part Number: MUX36S16
Other Parts Discussed in Thread: TMUX136

Hello E2E!

I was looking at the MUX36S16 and I'm curious if anyone has tried to use this for the purpose of switching between USB ports using two MUX36S16's (for positive and negative signals respectively) or a solution of the sort? I have a need to switch one USB host port to one of 16 devices programmatically and I'm wondering if this could be a potential solution.

Edit: This is for USB 2.0 full-speed (480Mbps max)

Thanks,

Oliver

  • Hey Oliver, 

    Truthfully this isn't something i've seen be done before. I can't say i've seen many use cases that require 16 USB2.0 signals to be muxed down to one. You may run into some problems with regards to the frequency range. Could you clarify more on the data rate. Is that 480Mb per line? I'm not terribly familiar with how the USB protocol is actually read so how the 480Mbps is achieved will impact the frequency range that you're looking for (for instance, is there only one read per cycle?). 

    That being said, iIm a little more skeptical of the frequency listed since it isn't actually on the datasheet. If you know the loading conditions (RL an CL), we can approximate this. I would still be skeptical if the speeds can be achieved since we typically recommend to have 1.5-3x the frequency of the data being passed through as the bandwidth on your switch since it's really quite load dependent. 
    Truthfully you may have this problem in general for a 16:1 mux. Have you considered a USB hub? If this is an approach you're interested in, you could either post in the data converter forum, or I can loop them into this thread and we could come up with a solution that may involve a couple hubs and a multiplexer. I'm not so confident in a 16:1 solution with that speed though.

    Thanks,
    Rami   

  • Hi Rami

    Thank you for taking the time to look into this. I believe an explanation of the use case of the circuit would give a better idea of how it's supposed to work: A USB master device will connect to one (only one) of 16 slaves at any given point to transfer data cross. Both the master and slave devices only support USB 2.0, hence the max theoretical bandwidth of 480Mbps. I was looking at the MAX4999 IC for this, as it nearly matches my needs, but "only" does 8:1 (quite singificant in this scenario, still) with 1200MHz@-3dB. The circuit itself is still very much theoretical, I'm only making my thoughts about how to design it at the moment and was hoping someone would have a solution close to the MAX4999, but obviously with the more circuits. I think a USB hub and two MAX4999's could be my best option in this scenario, though.

    Thanks again,

    Oliver

  • Hey Oliver,

    Just adding another level of thinking here and I may be misunderstanding but I believe you can narrow down the configuration if you end up using a hub. That's to say if you need 16 peripherals in the end, and you find a hub with 8 channels then you could in theory just mux the original controller signal between the two hubs with a 2 channel 2:1. The TMUX136 would be a good choice here. 

    Thanks,
    Rami

  • Hey Rami

    I do truly appreciate the input, but I think it's better in this scenario to do the opposite and use a MAX4999. So put one 2-channel 2:1 MUX in front of two MAX4999's. As the downstream peripherals would benefit from using the full USB bandwidth, it is in my interest to only have one USB peripheral active at a time. Perhaps using a TMUX136 and putting the A and B channels towards a respective MAX4999 to go further downstream. Using a USB hub instead of the TMUX136 in this setup is surely possible, but given only one peripheral needs to be active, I don't think there's any benefit to it, other than not having to control which port is active. But that is the least of my concerns at the moment.

    Thank you for bringing these considerations into the fray. I think I've got my solution now!