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.

PLC SOM with 3-phase support (both hardware and firmware)

Hi there,

We are building a product with 3-phase PLC prime support, and find the following two solutions fits our needs pretty well:

Data Concentrator Cape for BeagleBone Black ( www.ti.com/.../TIDA-00225 )
Data Concentrator Reference Design ( www.ti.com/.../TIDEP0006 )

However, in the technical documents, "PLC SOM" were referred but no model number was mentioned. We thought it was SOMPLC-F28PLC83, but was confirmed by TI that F28PLC83 only supports single-phase.

So our question is, is there any PLC SOM supporting 3-phase available now? If yes, what the model number and where we could get it? If not, is there any reference design (even conceptual one is fine) available?

Thanks

Henry

  • Hi Henry,

    What do you mean by 3-phase PLC PRIME support? You mean sending and receiving from 3-phase lines? Basically, The PLC on F28PLC83 and F28PLC84 has the same set of features and can support 3-phase communication by wiring to 3-phase power-line. Please, refer to the data concentrator reference design for this. For the SOM for the design, both F28PLC83 and F28PLC84 should work and software is available on

    .

    Regards,

    Wonsoo

  • Hello Wonsoo,

    Thanks for your reply.

    Yes, we would like to have 3-phase PLC support to send and transmit messages along all 3-phase AC lines. In our application, the PLC Prime meters could be connected to any of the three phases lines.  

    What confuses us is that in your reference design:  Data Concentrator Reference Design ( http://www.ti.com/tool/TIDEP0006 ), PLC SOM is used as part of the design but no PLC SOM model # is mentioned, Only information we could find regarding this PLC SOM is the pin out shown in sheet 26 of the schematics. I have pasted part of the schematics here.

    as you can see from the PLC SOM's pin out, our understanding is that to connect three-phase to PLC SOM, ZERO_PLC_A, ZERO_PLC_B and ZERO_PLC_C should be connected to pin-10, pin-30 and pin-19 respectively.

    However, if you look at F28PLC83's design documents ( www.ti.com/.../TIDM-SOMPLC-F28PLC83 Documents ), pin-10 is ZC which looks like ZER_CROSS signals from a single phase, but pin-30 is GPIO04 (Unused multi-purpose IO pin, GPIO04), and pin-19 is GPIO26 (Unused multi-purpose IO, GPIO26). From this document, we cannot see how to connect this F28PLC83 to allow it support three-phase PLC Prime. 

    If you are aware of any reference design using F28PLC83 to support 3-phase PLC prime, would you like to share it with us?

    Thanks a lot

    Henry

  • Hi Henry,

    Here are some comments.

    1. F28PLC83/F28PLC84 devices are the same pin configuration and as you mentioned both devices support single-phase ZC detection via pin-10 (ZC: GPIO9). Please, do not connect the other ZCs in the reference schematic because the GPIOs are used in different purpose now.

    2. Supporting single-phase ZC detection only doesn't mean that the PLC can't transmit over 3-phase lines. As shown in the data concentrator reference design, wiring three 3-phase lines after the transformer can achieve to transmit and receive over 3-phase lines.

    Thanks.

    Wonsoo

     

  • Hello Wonsoo,

    Thank you for your prompt reply. From data concentrator reference schematics, it appears that zero crossing signals for three AC phases are generated separately (as shown in sheet 25, signal ZERO_PLC_A, ZERO_PLC_B and ZERO_PLC_C), before three AC lines are coupled together to form PL_TXRX (as shown in sheet 25).

    Since the zero crossing signal is used to synchronizing PLC prime transmitting, without proper zero-cross signal, how to ensure messages are properly synchronized to each phase, especially during transmission?

    Maybe we are missing something here? More suggestions are highly appreciated.

    Thanks

    Henry
  • Henry,

    Transmission is not necessarily synchronized with ZC. Based on PRIME standard (or most of OFDM-based PLC standards), the synchronization at receiver side is achieved by preamble detection not by ZC detection. I know some FSK-based systems are synchronized with ZC signal but this is not the case.

    Regards,
    Wonsoo
  • thanks Wonsoo, this clarifies our concerns. We will give F28PLC83 a try to see how if works with three phases. If we encounter any problem, we might still need your help...

    Thanks a lot

    Henry
  • Hello Wonsoo,

    Sorry for the long delay and we are switching back to this project again..

    In our DC (data concentrator) application, we plan to talk to TI PLC module SOMPLC-F28PLC84 based on CCSC-4-32 (follow TI document PRIME DC IEC-61334-4-32 API Specification V0.1).  On the hardware side, we will couple all three phases together, and have one Zero-crossing input (ZC) connected to PIN-10 of the 34-pin header of SOMPLC-F28PLC84. 

    For transmitting, the DC will send out message to all three phases. For receiving, devices (Service Node) on any phase could respond to DC's (Base Node) requests. 

    The question we have now is:  when we received a response message, how to know from which phase (A, B or C) the response comes, because all three-phase AC are coupled together, and only have one ZC input. We checked PLC prime Spec V1.4, section 3.11.3.13, it specified that "The primitive is generated by the MAC layer when it needs to know the zero cross time of the mains, e.g. in order to calculate the phase to which the Node is connected", but in TI's 61334-4-32 API documents, looks like there is not PHY_ZCT information available, maybe just because it is in MAC or PHY layer therefore we don't have access to it.

    Appreciate if you could advice how to do phase detection for incoming message when using F28PLC84  module.

    Many thanks

    Henry

  • Hi there, can we get any technical support on this matter? Thanks, Henry
  • Henry,

    Sorry for late response. 

    We've created a mechanism to detect absolute phase detection with a single ZC, which is available in the latest PRIME release on the TI website. Here is the idea. 

    1. The DC enables a single ZC detection and assume that the DC knows which phase the DC is attached. 

    2. Any nodes (including DC) can get the distance between beacon (seq #0) and ZC time locally (via PIB access).

    Please, check out PRIME SN software () package and refer to page 20 in PRIME_Host_Message_Specifications.pdf. 

    The related PIBs are:

    Distance between beacon and ZCT

    Zero-crossing time period

    3. Deliver the information to the DC (via application data) and so DC can know which phase each node can attach to by comparing the DC's distance vs. each node's distance. The zero-crossing time period (50 or 60 Hz) can be reference to calculate the phase distance in time. 

    4. Note that the distance is updated based on beacon sequence number 0 and so the delivery should be done within superframe time (32*640ms). In addition, to use this mechanism, you need to use TI modem for both SN and DC. 

    Thanks.

    Wonsoo

     

  • Hello Wonsoo,

    Thank you for your detailed explanation. We will explore your suggestions a bit more, and see how it works.

    Best Regards

    Henry
  • Hello Wonsoo,

    Couple of more questions following your suggestions:

    1) Can we get those PIBs (Distance between beacon and ZCT, Zero-crossing time period) from SN through the management interface of TI prime DC module?

    2) Can we try the feature using the DC software module on TI's website?

    http://www.ti.com/tool/ti-plc-prime-dc?keyMatch=prime%20dc&tisearch=Search-EN-Everything

    Thanks a lot

    Henry

  • 1) No. I guess that you are asking remote PIB access. The SN's PIB should be obtained via application level data delivery.
    2) Yes, the latest PRIME SN/DC releases include the feature.

    Thanks.
    Wonsoo
  • Hello Wonsoo,

    Thanks for your reply..Here are where we are based on your suggestions:

    Our settings are:

    DC (Base Node): our own AM335x based processor board (with "pdc_mgmt_cli" and "pdc" running) + F28PLC83 (BN)
    Service Node: Windows PC (with "hostappemu.exe" running) + F28PLC83 (SN)

    With this setting we could get get-node-info, get-switch-info.. and etc working. and also we could use "get-pib" to get local (BN) PIB attribute values.

    The raw message we captured between pdc_mgmt_cli and pdc is as below:

    Request:
    0x59 0xc0 0x00 0x10 0x04 0xa1 0x00 0x0c
    0x06 0x26 0x00 0x02 0x00 0xf2 0x06 0x27
    0x00 0x02 0x00 0x00

    Comfirm:
    0x59 0x00 0x00 0x06
    0x06 0x20 0x00 0x02 0x00 0x00

    Response (event)
    0x55 0x40 0x00 0x3f 0x09 0x00 0x00 0x3b
    0x09 0x01 0x00 0x02 0x01 0x13 0x09 0x02
    0x00 0x04 0x00 0x00 0x00 0x06 0x06 0x22
    0x00 0x19 0x46 0x72 0x69 0x20 0x41 0x75
    0x67 0x20 0x20 0x35 0x20 0x31 0x31 0x3a
    0x30 0x35 0x3a 0x30 0x30 0x20 0x32 0x30
    0x31 0x36 0x00 0x09 0x03 0x00 0x0c 0x06
    0x26 0x00 0x02 0x00 0xf2 0x06 0x21 0x00
    0x02 0x2f 0x0e

    However, when we try to send "get-pib -e 00:00:00:00:00:03" to get remote PIB from SN which is already connected, we could not receive any response, The Raw message we captured looks like this:

    Request:
    0x59 0xc0 0x00 0x1a 0x04 0xa1 0x00 0x16
    0x06 0x26 0x00 0x02 0x00 0xf2 0x06 0x27
    0x00 0x02 0x00 0x00 0x06 0x03 0x00 0x06
    0x00 0x00 0x00 0x00 0x00 0x03

    Comfirm:
    0x59 0x00 0x00 0x06
    0x06 0x20 0x00 0x02 0x00 0x00

    Response
    (NULL)


    We really appreciate if you could help us to see what could be the reason that we could not received SN PIB. According to TI Document "TI PRIME DC Software User Guide Ver 1.6" section 3.1.27 (page 20), "get-pib" command with -e parameter should be able to get the SN (remote) PIB attribute values.

    From the same document, it appears the private PHY PIB 0xF1 and 0xF2 were added to get zero-crossing timestamp to be used for "phase-detection". Are these two PIB supported by a normal/standard PLC Prime meter? In another word, if the SN is a third party meter, can we use the same sets of PIB to achieve "phase detection"? if not, is there any alternatives defined by standard PLC Prime?

    Thanks a lot

    Henry