• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 34 subscribers
  • Views 277 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.

SN65HVD233: minimum load resistance

Brian Angiel
Expert 3345 points
Part Number: SN65HVD233

Hi team,

I have a question regarding a SN65HVD233. I’m trying to determine the minimum load resistance this device is capable of driving. All I can find clearly stated in the datasheet is 60ohm+/-1% for various parameters, which I presume may be a test condition and not an operating range.

That said, I’m calculating the minimum load resistance based on drive capability. Is it correct to use the value shown in Table 8.3 (50mA) in combination with minimum Differential Output Voltage (1.5V)?

1.5V / 50mA = 30ohm

Thanks for any assistance.

Brian

  • 0
    TI__Mastermind 49370 points
    Hey Brian,

    I've assigned an engineer to this thread for you. You should expect a response by tomorrow.

    Thanks,
    -Bobby
  • 0 in reply to BOBBY
    TI__Intellectual 2090 points
    Hi Brian,

    Thank you for your question.

    Could you let me know what design challenge is driving your load resistance choice? I ask, because usually the load resistance is chosen by the characteristic impedance of the network cabling in the system. This is done to minimize noise and reflections in the network, and standard cabling typically possesses a characteristic impedance of 120 ohms. Placing one termination resistance on both ends of the bus line produces a typical effective bus load of 60-ohms. This is the reason why the datasheet focuses its performance at that loading level.
    I have seen designs in the past implement cabling with lesser characteristic impedance (about 100 ohms), bringing the loading resistance down to 50 ohms. Our CAN technology is able to support communication at this loading level as well.
    On another note, power consumption will also increase significantly with decreasing load resistance.

    Regarding your calculation above, I am cautious to take the 1.5V dominant measurement at 60 ohms and extrapolate it to a smaller resistance.
    We can take some measurements from our end to give you differential voltage and current at lower load resistances. 30 ohms is pretty low though, given what we normally see in the industry. Typically, we don't see load resistance drop below 45 ohms.

    Let me know why you see the need to drive lower load resistance, and we can get some measurements back to you for a recommendation.

    Best Regards,
    Max Megee
    TI TRX Applications
  • 0 in reply to Max Megee
    TI__Expert 3345 points
    Hi Max,

    My design will use an unshielded twisted pair with a characteristic impedance of 110ohm. I will have a minimum of 33 nodes with the potential for more. I’m budgeting for a total of 40 nodes. My data rate will be relatively slow at 250k, and the cable length will be relatively short (approximately 6 meters). Stubs will be minimal, perhaps 0.5 meters, absolute worst case.

    The load capability of SN65HVD233 is unclear to me. I realize nominal load is 60ohm, assuming only two nodes, that would be almost true. But I need to take into account the load impedance of the transceivers. At the worst case 40kohm, as stated on the datasheet, with 33 nodes total, 32 receiver loads plus the termination, equates to about 57ohm [(40k/32 * 60) / (40k/32 + 60)]. If the specification is to drive loads of 60ohm +/- 1%, 57ohm is outside the specified range.

    In short, I’m looking for some numbers in the datasheet to prove that this will work. There’s info in the datasheet describing how it work with 120 nodes, but I don’t see any specific numbers to indicate how this is achieved. Any details you can provide on this would be appreciated.

    Thanks,
    Brian
  • 0 in reply to Brian Angiel
    TI__Intellectual 2090 points

    Hi Brian,

    Thank you; I understand your design challenge now.  One thing to clarify, when considering the effect of the impedance of each receiver node in the system, the receiver's impedance should be treated as common-mode loading, not differential loading.  This is because the receiver itself biases the common-mode voltage to VCC/2.  Therefore, each node introduced will add parallel common-mode resistance to system, resulting in a circuit equivalent to the figure below:

    In the figure, the extreme example of 120 nodes is shown.  Taking 120 nodes with a typical single-ended receiver impedance of 40k ohms, gives an equivalent circuit of two 333 common-mode resistors biased to the V_TEST voltage.  

    In this case, the differential load remains 60 ohms across the driver.  Again, this is due to the influence of the internal biasing of each receiver to VCC/2.  The V_TEST voltage range in the figure is given in order to characterize the device across possible ground shifts in the system.  

    Therefore, in your design, the challenge becomes one involving robust network design with minimal stub lengths to reduce reflections as much as possible.  With good network design, as you have outlined above, keeping stubs <0.5m at the total end-to-end length to 6m, then I think you have a robust design that should support 40 nodes.  

    Does this make sense?  Anything I can clarify in the impedance analysis?

    Best Regards,
    Max