• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 16 subscribers
  • Views 552 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Original question:

SN74CBTD3861: Voltage translation with and without FPGA logic loaded

SN74CBTD3861: More complete recommendations for using SN74CBTD3861 as voltage translator between FPGA and 5V logic

Charles Gardiner
Prodigy 60 points
Part Number: SN74CBTD3861
Other Parts Discussed in Thread: SN74CBT3861

Hej,

we are considering using the SN74CBTD3861 for interfacing between a 3.3V FPGA and 5V external logic.

I have looked at the following link and the footnotes in the data sheet but am still not completely clear about the optimal connect for the SN74CBTD3861 device: https://e2e.ti.com/support/switches-multiplexers-group/switches-multiplexers/f/switches-multiplexers-forum/592085/sn74cbtd3861-voltage-translation-with-and-without-fpga-logic-loaded/2175708?tisearch=e2e-sitesearch&keymatch=SN74CBTD3861#2175708

1) The forum item above mentions that 5V appears on the FPGA side if the FPGA is not yet configured. Does this mean we should really have the gate switch OEn turned off until the FPGA is loaded to ensure that the FPGA never sees 5V?

2) There is a note in the data sheet, "All unused control inputs of the device must be held at VCC or GND to ensure proper device operation." and a reference to application note SCBA004. SCBA004 seems more of a general discussion on floating CMOS inputs. In our application we will be using individual CMOS switches in both directions, sometimes possible even different directions in the same package. i.e. ometimes the FPGA is source, sometimes the FPGA is consumer/sink. So how should we tie an unused CMOS switch pair? FPGA side to ground? FPGA side to 3.3V? FPGA side to 5V (because not connected to FPGA/unused)? Logic side to ground? Logic side to 5V?

3) Any other recommendations are of course welcome.

Thanks.

  • 0
    TI__Mastermind 35648 points

    Hey Charles,

    1) This would be good practice to isolate the paths until things are configured. However, in the E2E post you shared, it seems that the 5V comes from a pull up on the FPGA side, not the mux. You shouldn't see 5V on the I/O. That being said, is the objective here to just voltage translate down the 5V to 3.3V bidirectionally? A voltage translator may be a better solution

    2) This is referring to the logic control pins. So in this device's case it would just be the /OE pin. Tie this high (5V) or low (0V) when it is unused. However, i'd assume this will be actively driven high or low through the operation. 

    3) The recommendation I have here was what I sort of touched on in number 1. It seems you're looking to translate down a 5V I/O to a 3.3V O/I (bi-directionally). Let me know if this is what you're looking for and I can loop in the voltage translation team to help find the correct part for you. In the meantime you can also look through the product folder for a device that may be of use ( https://www.ti.com/logic-voltage-translation/voltage-translators-level-shifters/products.html )

    Thanks,
    Rami 

  • 0 in reply to Rami Mooti1
    Prodigy 60 points

    Hej Rami,

    many thanks for the really fast reply.

    Your assumption under 3 is correct. The FPGA is really just a bunch of registers with 32 inputs and 16 outputs. Depending on the way the board routing etc works out, my customer would like to be able mix and match inputs/outputs in the same package. FPGA is 3.3V, the input/output relays, sensors etc are 5V.

    This is a redesign and as far as I am aware the customer is currently using Renesas IDTQS3861. This is quite an old board, part of an industrial application which is still alive and kicking. It may well be that the Renesas Quickswitch product was not the ideal solution at the time. In any case, the application is 3.3.V FPGA, 5V peripherals and switching speed max. 20 MByte/s, probably well below that in practice.

    A footnote though, the customer may actually require pin compatibility to the Renesas Quickswitch. I'm not sure whether he is actually going to redesign the board.

    Thanks.

    Charles

  • 0 in reply to Charles Gardiner
    TI__Mastermind 35648 points

    Hey Charles, 

    Thanks for more information on this application. The Sn74CBT3861 would be a drop in replacement for the IDTQS3861 as it seems that's what you're aiming to replace here. Let me know if you have any further questions.

    Thanks,
    Rami

  • 0 in reply to Rami Mooti1
    Prodigy 60 points

    Hej Rami,

    thanks again for your reply.

    Just to make sure I got this right, you are recommending the Sn74CBT3861 instead of the Sn74CBTD3861? The data sheet for  Sn74CBT3861 doesn't mention voltage translation, the Sn74CBTD3861 does. The IDTQS3861  also performs voltage translation.

    Thanks.

  • +1 in reply to Charles Gardiner
    TI__Mastermind 35648 points

    Hey Charles,

    Oops, sorry on that one! Looks like I had a small typo and ommited the 'D'. The CBTD device is the recommended device, with voltage translation. Sorry about that and thanks for doubling back to confirm

    Thanks,

    Rami