TJA1410 Support - Does not accepting commands Hi Team, I am using the TJA1410 as the External Transceiver for our Project (using the TJA1410A-EVB). Our SoC (which having integrated T1S dPHY) is sending some commands to the Analog PHY, but the commands are not accepting by the TJA1410A. I have few doubts regarding that: 1. Whether the signals sent to the TJA1410 should need to obey the exact timings whatever mentioned in the OA3P document (OPEN_Alliance_10BASE-T1S_PMD_Transceiver_Interface_1p5_Final.pdf), what if we are sending the different Duration clock pulse to the External Analog PHY, will it work still? Which means the clock Frequency for the 10baseT1S-dPHY IP should need to be a fixed one, is it necessary? 2. I am supplying 3.3V to the EVB for VIO and VCC and having the 1.8V (LOGIC HIGH) to 0V (LOGIC LOW) for the RX/TX/ED Inputs, whether the TJA1410A silicon has any Internal Voltage Level shifters? I can't be able to Find the datasheet in the page. 3. Can I get any App notes for this TJA1410A Transceiver? We are in the mid of our Project so can you please consider this as High-Priority case and reply ASAP? Thanks, Madhan Re: TJA1410 Support - Does not accepting commands Hello @Madhan1 ,
The most likely issue is the voltage-level mismatch between your SoC and the TJA1410A.
OA 3-pin interface timing
The OPEN Alliance 3-pin interface is clock-less, so there is no separate interface clock with one fixed frequency. However, the TX command and data pulse widths must meet the timing requirements defined by the OPEN Alliance specification and the TJA1410 data sheet.
If the pulse durations are outside the specified ranges, the TJA1410 may not recognize the command correctly. Therefore, please verify the TX waveform directly at the TJA1410A-EVB connector using an oscilloscope or logic analyzer, including the pulse widths, signal rise and fall times, and the command sequence.
VIO and digital signal levels
The TJA1410A mounted on the TJA1410A-EVB is the 3.3 V VIO variant. When VIO is supplied with 3.3 V, a 1.8 V HIGH level on TX is not sufficient to meet the guaranteed input HIGH threshold.
The device does not provide internal level shifting that would allow a 1.8 V host to directly drive the TJA1410A while VIO is supplied with 3.3 V.
For a 1.8 V digital interface, please use one of the following options:
use the TJA1410B variant, which supports a 1.8 V VIO interface, or
add a suitable external voltage-level translator between the SoC and the TJA1410A.
Please also note that RX and ED are driven according to the selected VIO voltage, so the voltage compatibility must be checked in both directions.
Documentation
The relevant documents are:
TJA1410 data sheet
AN14787, 10BASE-T1S Ethernet PMD Transceiver TJA1410
UM12432, TJA1410-EVB Evaluation Board User Manual
OPEN Alliance 10BASE-T1S PMD Transceiver Interface Specification, version 1.5
Some TJA1410 documentation may require an NXP account, NDA and Secure Access approval.
As the first debugging step, I recommend correcting the 1.8 V versus 3.3 V interface mismatch and then checking the TX command waveform against the timing limits in the TJA1410 data sheet.
Best regards,
Pavel Re: TJA1410 Support - Does not accepting commands Hi @PavelL , Thanks for your timely Reply. And I have confirmed that in our setup we are having the External Level Shifter for converting the 1.8V to 3.3V for the TX/RX/ED Pins. Meanwhile what is the purpose of Internal Level Shifter present inside the TJA1410A Transceiver? I initially thought the Internal Level shifter will take care of this Voltage conversion. Anyhow we are having the External Level shifter so no worries. And our VCC and VIO supply is also 3.3V. And the termination Resistors also connected with the Headers. Regarding the OA3P timing specs, for example if I am sending the CONFIG command to TJA1410A as 80ns Low/80ns High/64us Low instead of 20ns Low/20ns High/16us Low will the TJA1410A will recognize it as CONFIG command or RESET command (because as per Ideal OA3P spec 80ns is for the RESET command)? Thanks, Madhan Re: TJA1410 Support - Does not accepting commands Hello @Madhan1 ,
If an external bidirectional level translator is already used between the 1.8 V SoC interface and the 3.3 V TX/RX/ED interface, the voltage domains can be compatible.
The “Transmitter Level Shifter” shown in the TJA1410 block diagram is part of the internal analog transmitter path. It converts internal control signals to the voltage domain required by the analog MDI transmitter. It is not a level translator for the external TX/RX/ED host interface.
Regarding your CONFIG command example, the proposed timing:
80 ns LOW
80 ns HIGH
64 µs LOW
is not a valid CONFIG command timing for the TJA1410.
According to the TJA1410 detection limits published in the Data Sheet, Table 14 :
the CONFIG LOW pulse must be between 10 ns and 33 ns,
the CONFIG HIGH pulse must be between 10 ns and 40 ns,
the following CONFIG assertion time must be at least 15 µs.
An 80 ns LOW pulse is outside the CONFIG detection range. Moreover, 80 ns falls within the RESET command detection range of 55 ns to 120 ns. Therefore, the first 80 ns LOW pulse may be interpreted as a RESET command after TX returns HIGH.
Best regards,
Pavel Re: TJA1410 Support - Does not accepting commands Hi @PavelL , Thanks for your reply. I understood your opinion regarding the timings. One more doubt, my hardware setup is like I am having the Pull Up Resistors of 47Kohm connected in all the three pins (RX/TX/ED) does this cause any issue in the communication and the External Pull Up resistors are recommended? I have connected only two nodes one side is my SoC and other side is the Linux PC (connected via 10baseT1S to USB bridge). I think we don't need to worry about the above node details because currently we are stuck with the Analog PHY wake up itself, if it wakes up then only, we can proceed further by transferring some packets. Thanks, Madhan Re: TJA1410 Support - Does not accepting commands Hello @Madhan1 ,
External 47 kΩ pull-up resistors on TX, RX, and ED are generally not required because the TJA1410 already provides the relevant internal pull-ups.
Due to their relatively high resistance, the 47 kΩ pull-ups are unlikely to be the primary cause of the issue. However, they may affect the signal edges together with the level translator and line capacitance, and RX and ED can change direction during Configuration mode.
For debugging, I recommend temporarily removing the external pull-ups and measuring TX directly at the TJA1410 pin. Please verify that the RESET command meets the specified timing and check whether RX becomes HIGH and ED LOW after the transition to Normal mode.
Bets regards,
Pavel Re: TJA1410 Support - Does not accepting commands Hi @PavelL , Thanks for your support. Using your answers we were looking into the issue, I think we have almost figured out, if any more doubts, I will ping you back in this same thread. Thanks, Madhan
記事全体を表示
