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.

programmable xbit buffer transceiver

Other Parts Discussed in Thread: SN74LVC125A, SN74LV126A, SN74LVC126A, SN74LV125A, SN74AVC8T245, SN74LVC8T245, SN74AVC4T774

Hello!

I wanted to ask if there exists a logic device that has xbits (as much as possible) to act as a logic buffer which I can program the direction via binary code or serial interface?

Basically the idea would be to have a device with 16 bits and 4 address bits which tell the device if bitx is an output buffer or an input buffer (driving about 10mA - 24mA). This device would have the advantage that i can create a default main host controller board where I buffer all available GPIOs and later decide - depending on the actual application - which ports I want to define as an input or output but with the advantage to have all the ports buffered allready.

All available transceivers I have found only have the possibility to change the input/output behaviour for all the bits of one element at a time.

I would be really happy if there exists a device like i've wished for.

best regards

Benjamin

  • Hi Benjamin,
    I don't know of a device that would do what you're describing.

    Can you tell me what the application would be?
  • This is not a common application, so there is no single device for this.

    Combine multiple buffers with opposite OE signals (e.g., SN74LV125A/SN74LV126A or SN74LVC125A/SN74LVC126A). If there are too many parallel OE signals, add shift registers.

  • I already described the application: I have to build up a kind auf default MC-Board which I can use for different applications. I just need the GPIOs of the MC to be buffered for further distribution..
  • Thanks!
    This seems to be a possible solution. I will look deeper in to it.

    best regards!
  • Thanks Benjamin,
    From your original description it didn't make sense why you would need buffers since GPIOs are already buffers that are direction selectable. Your application is to add current drive strength. Are you also protecting the inputs to the MCU? Or are the input buffers just 'along for the ride?'

    Have you considered also adding voltage translation to this setup? Having the ability to communicate to any voltage you like would be ideal. I've seen people do this before with banks of 1, 2, 4, and 8-bit direction controlled translators like the SN74AVC8T245 (for lower voltage applications) or SN74LVC8T245 (for up to 5V applications).
  • Sorry, I thought it was clear that I need the current drive strength, off course almost any microcontroller has the capability to drive something from 2 -4mA, I think. Thanks for your hint with the level translation, thats also a good idea.
    At least for the STM32s I have input protection for the ports.

    I don't know why no vendor is coming up with some programmable n-bit buffer IC , where the ports direction can be programmed via serial or BCD. In my opinion this would be a great deal, since it would be possible to create a digital board suitable for different products on a modular basis.
  • Hey,

    i just have one little question if my thoughts are correct or not:

    Let's take a eight GPIO (8-bit) example. I would have 8 programmable GPIOs which I want to have universally acces as Input / Output or a alternative signal (SPI or I2C, USART...). I would use a 8:16 demultiplexer, so that I can feed the following line transceivers with level translation. Either demultplexer output 1 for the transceiver programmed as output and output 2 for the transceiver programmed as input. I could then later decide which GPIO acts as an Input or Output and have a good current drive strength. Off course I shuld check propagation delay to stay as low as possible when choosing the right chips for that purpose. Correct?

    best regards

    Benjamin

  • Have you looked at how the internals of an MCU work? You're basically recreating that circuit outside the MCU, so it might be good to use that as a blueprint.

    You can using direction controlled level translators like the SN74AVC4T774 or the SN74LVC8T245 instead of normal transceivers rather than in addition to them.

    In my opinion, propagation delay shouldn't really affect your interfaces since they likely won't be high speed communication buses anyways.
  • Yeah you are right. The demultiplexer or 2:1 switch mode will not work, since I can only switch n-bits with the select signals not only a single one, or I would need to have a lot of single channel packages...damn..
  • Hey Emrys,
    I think I have a possible solution I would try: I will use two of the SN74LVC8T245. One configured as input and the other one as output. I will then route each port to both devices with a series resistance as option. I then can later decide if the port is an input or output with the applied o Ohm resistance. There is one problem that might be killing my plan: On the B side of the device I would also have to connect the B1 of the input and B1 of the output SN74LVC8T245. Can this be a problem when the SN74LVC8T245 gets a high or low signal on the B side configured as output?

    Mhh..when I think about it, I fear this will not work.

    bes tregards

    Benjamin