• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 18 subscribers
  • Views 2265 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TXB0108: Reading TTL levels from voltage divider

Xavier Saura
Prodigy 30 points
Part Number: TXB0108
Other Parts Discussed in Thread: SN74LV1T34, SN74LVC1G17

I’m trying to use the voltage translator TXB0108 in order to activate a relay (green dotted box) with a microcontroller through this voltage translator (blue dotted box). The connection between the two regions is done using flat IDC cable.

The circuit in the green box works really fine by itself, and also the circuit in the blue zone works good. Perhaps, I have some problems when interfacing both circuits, which we though are related to impedance mismatching and/or due to capacitance of the IDC cable.

Our idea is to activate a relay and afterwards, we would like to read its state through voltage divider (R2 and R4).

Well, when interfacing the circuit, we could activate the relay, perhaps we are not capable to switch correctly the “ACTIVATION FROM MICROCONTROLLER” node, but we can’t read the high value in “MICROCONTROLLER READING” node.

We did several tests with capacitors in series/parallel with R2 and/or R4 but we didn’t find the solution.

Maybe a solution could be using TXS0108 but we are not sure.

It would be really helpful to receive some help. Thanks in advance.

  • 0
    TI__Guru* 96591 points

    Hi Xavier,

    The TXB0108 has very limited output current - we rate VOHA at 20μA.  It's likely that your circuit requires more current than that.

    Since both translation stages appear to be unidirectional, I would recommend using a buffer instead. The SN74LV1T34 would work for both directions. Translating from 3.3V to 5V you would set VCC to 5V, and translating down to 3.3V you would set VCC to 3.3V.

    If you see slow or noisy edges at the 'READING OUTPUT' node, I would recommend using a schmitt-trigger device there (such as SN74LVC1G17).

  • 0 in reply to Emrys Maier
    Prodigy 30 points

    Hello Emrys,

    Thanks for your answer, we thought  that it was because of impedance mismatching but we were not sure.

    Regarding your proposal, we would like to use a bidirectional voltage translation if possible. In additon, we saw that your proposed solutions are single channel. Maybe it would be better to use 8-channel solutions and much better if it has the same pinout.

  • 0 in reply to Xavier Saura
    TI__Guru* 96591 points

    I haven't seen a reason to use bidirectional translation or 8 channels, but if that's what you insist on - I might recommend trying the TXS or LSF series of translators.  They can provide more current than TXB and come in 8 channel options, however they are passive translators and have their own quirks.  You can check out these application reports for details on how each works:

  • 0 in reply to Emrys Maier
    Prodigy 30 points
    It's a more complex design but I reduced it in order to clarify the question. The main reasons to use a bidirectional one are to reduce the numbers of ICs and of course the space in the design. We have 3 similar voltage divider readings and 5 digital outputs (3 of them similar, and 3 for MOSFET activations) that we would like to join in a single voltage translator.
    We tryied to use TXS0108 but it is still not enough. We need to use a transistor and a pull-up between the node "reading output" and the TXS pin. We think that it's because of mismatching impedances.
    Now, we will try with your recommendations, thank you very much!!
  • 0 in reply to Xavier Saura
    TI__Guru* 96591 points
    Thanks for the clarification Xavier.

    Instead of using a single translator, you could use a pair of hex or octal buffers. I know it could cause some routing difficulties, but it shouldn't increase board space requirements and would be a much better solution than the bidirectional translators, at least in my opinion.