• State TI Thinks Resolved
  • Locked Locked
  • Replies 27 replies
  • Answers 6 answers
  • Subscribers 21 subscribers
  • Views 1390 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.

SN74GTL2003: SN74GTL2003 query

Shinu.V.Mathew
Genius 10370 points
Part Number: SN74GTL2003
Other Parts Discussed in Thread: LSF0108, TINA-TI, LSF0102

Hi Team,

My customer is using SN74GTL2003RKSR Level translator in my Design, I have query on Sref & Dref Voltages. 

Applying Sref=3.3V, and Dref=Gref=1.8V to 5.5V,

Will it work fine for Dref varying from 1.8V to 5.5V ?.

Please find attached Image.

Thanks Shinu. 

  • 0
    Guru 316690 points

    The constraints of this device are:

    • SREF must not be higher than any signal voltage.
    • DREF must be at least 1.5 V higher than SREF.
    • GREF must be connected to DREF, and there must be a current-limiting 200 kΩ resistor between those and the power supply.

    The circuit shown above violates all these constraints.

    Please note that there are no other constraints on the signal voltages (the high level of each output is determined only by the respective pull-up resistor), so you can simply connect SREF to a fixed 1.8 V, and DREF to any fixed voltage between 3.3 V and 5 V.

  • 0 in reply to Clemens Ladisch
    Intellectual 435 points

    HI Clemens,

    Thanks for info.

    Do you have any Level translator  to support A side=3.3V and B side=1.8V to 5.5V?.

    Regarding above query on  SN74GTL2003,

     case 1: Sref=3.3V and Dref=5.5V

    Case2: Sref=3.3V and Dref=3.3V(don't want to bypass the device)

    Case2: Derf=3.3V and Sref=1.8V(will use one more device)

    Please let me know all 3 cases will work fine?

    regards,

    tatarao K

    9059058519

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    Hello,

    I would recommend using the LSF0108 configured for single supply translation. We have a set of videos that explains the operation of this device in great detail here: Understanding the LSF family of bidirectional, multi-voltage level translators | TI.com Training Series

    Please note that the LSF0108, just like the SN74GTL2003, is a completely passive device - it will not redrive signals, so the drivers must have sufficient strength to operate the translator from all devices on the bus.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    HI,

    This LSF0108 will support 5.5V Supply voltage?

    Below screenshot from this datasheet stating that it will support 5V only.

    1st Page and 5th(6.3 Recommended Operating Conditions) stating that will support 5.0V only, but i need 5.5V support. Please clarify me.

    Stock is not available in digikey. how we can procure?

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    Hello,

    This device supports operation up to 7V, but it's recommended range is up to 5V -- in other words we have extensively tested and characterized to 5V, but the transistors can operate up to 7V without damage.

    The TI store currently has some units for LSF0108PWR (I saw 30 when I checked ti.com). I'm afraid I don't have any information regarding availability -- you could contact your local TI sales office to get more detailed information for ordering and pricing.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    HI EMRYS,

    Thanks for reply,

    Regarding LSF0108PWR level translator,

    Datasheet stating in different ways about Vref A & Vref B Voltages as below,

    Page No 1: Allows bidirectional voltage-level translation between, 1.2 V Left right arrow 1.8/2.5/3.3/5 V

    Page no 13:   Vref_B = Vref_A + 0.8

    page no 13: Vref_B and VI(EN) are recommended to be 1.0 V higher than Vref_A for best signal integrity.

    From above 3 points, which one i can consider?

    Shall i apply  Vref_A=1.2V and Vref_B=1.8V, 3.3V, 5.5V ?, will it work properly?

    regards,

    Tatarao K

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    The datasheets does not contradict itself here.

    tatarao kuramdasu said:
    Page No 1: Allows bidirectional voltage-level translation between, 1.2 V <->1.8/2.5/3.3/5 V

    The device can translate signals between 1.2V and 1.8V, however the supplies in that case will not be 1.2V and 1.8V because those are too close together (must be 0.8V apart). A single supply solution is one option to solve this, or you can just use supplies with more separation if they are available in the system (for example, 3.3V and 1.2V).

    tatarao kuramdasu said:
    Page no 13:   Vref_B = Vref_A + 0.8

    This value is from the minimum column of the application operating condition table, which indicates that is the lowest value that can be used for operation.

    tatarao kuramdasu said:
    page no 13: Vref_B and VI(EN) are recommended to be 1.0 V higher than Vref_A for best signal integrity.

    This value is recommended for "best signal integrity" -- it is not a requirement.

    I would highly recommend watching this video series to get a full understanding of how the device works, which will help you greatly with your design. The whole series is only about 20 minutes: The Logic Minute – Logic and translation | TI.com Training Series

    tatarao kuramdasu said:

    From above 3 points, which one i can consider?

    Shall i apply  Vref_A=1.2V and Vref_B=1.8V, 3.3V, 5.5V ?, will it work properly?

    I see two possible solutions for you

    (1) Use a single supply configuration, either as shown in this video

    (2) Use the 1.2V supply as a voltage clamp (connected directly to the EN pin), leaving Vref_A and Vref_B disconnected.

    The first option is the best solution if you have a 1.8V or larger supply available to create the required circuit, and the second option is the best if you only have a 1.2V stable supply in the system.  The second option will require pull-up resistors to be added to both sides of the device as the clamp voltage will be approximately 0.4V instead of the desired 1.2V.

  • 0 in reply to tatarao kuramdasu
    Intellectual 435 points

    HI,

    Shall i use LSF0108PWR Level translator as below?.

    Verf B =3.3V  <----->VerfA=1V  <------>Verf=18V to 5.5V(variable input)

    I will keep 2.2K pullup resistors for all signals, all sides.

    and Enable pin & Verf B short it and will connect to B side power rail through 200K resistor.

    regards,

    tatarao K

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    While it is theoretically possible to do what you're showing, This type of translator usually can't be cascaded as shown.

    The device doesn't have any output drive strength to drive the next translator stage - it is completely passive. All of your drive strength comes from the devices connected to the LSF and the pull-up resistors.

    I can provide you with a Tina-TI simulation model that you can test your ideas with, but you will need to add appropriate strength drivers and parasitic capacitance for each channel to see the expected effects.

    5165.LSF0108_Example.TSC

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    ,

    HI EMRYS,

    Thanks for prompt response.

    Could you please suggest me suitable Level translator for my application?.

    requirements are,

    1. Should support UART= 1MHz, I2C= 3.34MHz and SPI =20MHz, Bidirectional translator.

    2. A side 3.3V and B side 1.8V to 5.5V variable voltage.

    it is an urgent and bulk requirement, please support ASAP

    regards,

    Tatarao K

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points
    tatarao kuramdasu said:

    Could you please suggest me suitable Level translator for my application?.

    requirements are,

    1. Should support UART= 1MHz, I2C= 3.34MHz and SPI =20MHz, Bidirectional translator.

    2. A side 3.3V and B side 1.8V to 5.5V variable voltage.

    LSF0108 using the single supply configuration would work for that. Please review the information I have already provided.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    HI EMRYS,

    Thanks for suggestion, i saw that video for single supply voltage.

    Shall i give Single supply for LSF0108PW also, or it is applicable for only LSF0102?.

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    All of the LSF family translators work the same -- you can use the same biasing techniques for all of them.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    I can use LSF0108PW for UART= 1MHz, I2C= 3.34MHz and SPI =20MHz signals by directional translation, right?

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    It will depend on your system characteristics (drive strength and parasitic capacitance).

    The video on up-translation includes the method I use to calculate pull-up resistor values and determine if a system _can_ handle the desired operating conditions.

    Screenshot from ~5 min mark:

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    1. What is the common resistor value i can use for 1.8V to 5.5V range

    2. Do we need to place 15pF cap for all signals to GND?

    3. 200K need use for ASF108 also

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points
    tatarao kuramdasu said:
    1. What is the common resistor value i can use for 1.8V to 5.5V range

    This is dependent on your system -- you must select resistors that will meet your drive capabilities and parasitic capacitances.

    tatarao kuramdasu said:
    2. Do we need to place 15pF cap for all signals to GND?

    Do _not_ add capacitors. The ones you see in the screenshot are representative of parasitic capacitance, which is stated clearly in the video.

    tatarao kuramdasu said:
    3. 200K need use for ASF108 also

    The 200k resistor is part of the bias circuit. Please refer to the bias circuit video for details.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    Do you have any method to calculate parasitic capacitance?

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    First you should add all devices connected to the I/O pin (including the LSF device itself, which adds about 15 pF).

    If you don't know - add 5pF per device.

    Next, the capacitance for a trace is typically calculated using the parallel-plate capacitance equation:

    C = ε0εrA/d

    ε0 is defined as 8.8541878128(13)×10−12 F⋅m−1

    εr is defined by the material of the capacitor (ie your board's material... often FR4 is used and that has a relative permittivity of ~4.1

    A is the area of the trace (length * width)

    d is the distance between the trace and ground plane (if you don't have a ground plane, this equation won't work for the calculation)

    Or you can use one of the many calculators available on the internet for trace impedance, or possibly even one in your board design software.

    These should all be summed up and the total used for the capacitive loading of the LSF device.  A typical value will range from 15 pF to 50 pF. Some systems will have more capacitance and probably won't work with LSF.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    HI,

    Please see below connections i made for LSF0108PWR, let me know if any suggestions

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    This looks correct to me.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    HI Emrys,

    For LSF0108PWR, i gave Verf B=3.3V and Vref A= 1.8V to 5.5V. 

    My Queries are,

    1. Please let me know if any suggestion on this connections,  

    2. Will it support I2C(3.4MHz), SPI(20MHz) and UART(1MG baud)?. because for these interfaces only i am using this Level translator.

    3. Also i Need DC-DC Buck-Boost converter, please let me know if any suitable part for below spec.

    Spec are, Input=2.7V to 30VDC, Output= variable or Fixed 5V @ 2.5A to 4Amp. 

    4. Do you have any USB to I2C(3.4MHz), SPI(20MHz) and UART(1MG Baud) Adapter?.

    regards,

    Tatarao K

    9059058519

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    (1) As I said above, the connections are correct.

    (2) 20 MHz is pretty fast for a passive translator -- you'll need a pretty low load capacitance (short traces) to ensure good signal integrity.

    From the up-translation video:

    Your pull-up resistors are 1.8k, and I'll assume 15pF loading for simplicity:

    f_max = Data Rate / 2 = 1/2 * 1/(6*1800*15pF) = 3.1 MHz

    To get up to  20 MHz you'll need smaller pull-up resistors or a smaller load capacitance -- the former being much easier. You need about 6.5x the speed, so a 6.5x smaller resistor is required. Using 277 ohms, the max frequency is:

    f_max = Data Rate / 2 = 1/2 * 1/(6*277*15pF) = 20 MHz

    A 277 ohm resistor will require a pretty significant amount of current to operate (18mA @ 5V supply), which could impact your VOL.

    These two factors (sink current and data rate) are what determine if you can use this translator.  It's up to you to determine this for your system.

    For items (3) and (4) -- I don't support those devices. Please create a new thread for each question and post to E2E so the questions can be split out ot the appropriate engineers for support.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    HI,

    Could you please suggest any suitable part which will support High speed.

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    The LSF family supports 20 MHz with a 15pF load and 277 ohm pull-up resistors, as I stated above.

    If you can remove the auto-bidirectional requirement, I have many devices that have much faster data rates - but for auto-direction sensing translators, there aren't many choices.

  • 0 in reply to Emrys Maier
    Intellectual 435 points

    HI Emrys,

    Thanks for response,

    1. As per datasheet we can go upto 15mA current, right?

    2. Do we need pullup resistors other then I2C signals also?.

    For I2C we should need pullup since open drain, but SPI and UART Signals also need to give pullup to LSF(A side and B side)?

  • 0 in reply to tatarao kuramdasu
    TI__Guru* 97046 points

    (1) The table provided is an easy reference, but does not limit the drive current -- this device does not provide drive current -- the amount of current is limited by your driver, not the LSF or GTL translator. 

    (2) Yes, pull-up resistors are absolutely required. Please refer to the up-translation video I previously referred to for details.