• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Subscribers 32 subscribers
  • Views 227 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.

THVD1450: Node and data rate

Lin_Charles
Expert 4260 points
Part Number: THVD1450

Hi team,

How many nodes can be conneted to bus when data rate is 5Msps using THVD1450. How to calculate the max nodes when data rate change? 

Thanks.

BR,

Charles Lin

  • 0
    TI__Guru 62920 points

    Hi Charles,

    There are a few different factors that can limit the number of nodes.

    One factor is the input leakage current of the transceiver.  The parallel combination of all nodes as well as any termination should typically result in an effective differential load of no less than 54 Ohms to align with the RS-485 standard.  The input leakage current of a transceiver is often specified in terms of "unit loading" which you can read more about here:

    https://e2e.ti.com/blogs_/b/analogwire/archive/2016/03/08/rs-485-basics-how-to-calculate-unit-loads-and-the-maximum-number-of-nodes-on-your-network

    THVD1450 is a "1/8 unit load" device, 256 nodes could be supported.

    Another factor is the total capacitive loading of the bus, which can distort higher-frequency signals and result in jitter.  For a recommendation on how to analyze this, you can refer to this post:

    https://e2e.ti.com/support/interface/f/138/t/773543

    The effective output impedance of the THVD1450 can be inferred through Figures 2 and 5 in the datasheet (depending on the VCC level used).

    A third factor relates to the cabling topology itself.  Often times connections are not made in an ideal linear "daisy-chain" fashion from node to node, and unterminated "stub" lines cannot be avoided.  Each of connections may generate reflections that degrade signal integrity.  This tends to be the limiting factor in most networks, but it is generally not a property of the transceiver (although the output transition times do influence how apparent the reflections may be for a given stub length).  You can read more about this phenomenon here:

    http://www.ti.com/lit/an/slla272c/slla272c.pdf

    Regards,
    Max

  • 0 in reply to Miguel Robertson
    TI__Expert 4260 points

    Hi Max,

    Thank you for your detail explanation here. I have explained this to customer and they appreciate your support here. But customer may need team's help to analyze below application for THVD1450,

    Data rate=5M,

    Cable length <1m;

    No other capacitance on the bus and 120ohm terminal resistor will be weld.

    What is the max node you can estimated in this? Thanks.

    BR,

    Charles Lin

  • 0 in reply to Lin_Charles
    TI__Guru 62920 points

    Hi Charles,

    At that data rate I think would be primarily limited either by the reflections from stub nodes or (if the network is wired linearly) the parasitic capacitance of each node and associated cabling.  We can try to collect some data next week showing these effects a little better to give you a good point of reference.  Which VCC do you plan to use (3.3 V or 5 V)?  And, do you have a target for how many nodes could be supported in this application?  Is the 1-m cable between adjacent nodes on the network, or is it the total cable length (longest distance between end nodes)?

    Thanks,
    Max

  • 0 in reply to Miguel Robertson
    TI__Expert 4260 points

    Hi Max,

    Customer need to support 32 nodes on the bus, VCC=3.3V, the 1m is the total cable length. Very appreciate your help here. Thanks.

    BR,

    Charles Lin

  • 0 in reply to Miguel Robertson
    TI__Expert 4260 points

    Hi Max,

    Do you have some result can be shared with customer? Thanks.

    BR,

    Charles Lin

  • 0 in reply to Lin_Charles
    TI__Guru 62920 points

    Not yet; we are still working to collect data.  If 1 m is the total cable length then I don't expect any issues supporting 5 Mbps communication among 32 nodes.  But we will still plan to provide more info once it is available.

    Max

  • 0 in reply to Miguel Robertson
    TI__Expert 4260 points

    Max,

    When can we get this result fixed? I think this is commen issue from customer, it would be great if you can share some documens for us to explain it to customer. Thanks.

    BR,

    Charles Lin

  • 0 in reply to Lin_Charles
    TI__Guru 62920 points

    Charles,

    Agreed that it would be nice to have the data.  We are planning to collect it next week.

    Max

  • 0 in reply to Miguel Robertson
    TI__Guru 62920 points

    Hi Charles,

    Our team has now had an opportunity to take some lab measurements related to this.  The differential capacitance of the THVD1450 is about 30 pF.  We tested the device with a constant 0/1/0/1 toggling pattern at 5 Mbps for an increasing capacitive load.  Here are the A (purple) and R (yellow) waveforms with no loading besides termination (i.e., 60 Ohms between A and B):

    Here are the same signals with additional 2.2 nF of loading:

    The bus rise/fall times are slowed as expected, but the R output still matches the data pattern.  This loading would be equivalent to ~73 THVD1450 units.  So, I think that this confirms that operation with 32 nodes should be achievable in most cases.

    Please let us know if you have further questions.

    Regards,
    Max