• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 35 subscribers
  • Views 294 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.

DS90UH948-Q1: Several signal related issues confirmed.

Alan Xia
Expert 8740 points
Part Number: DS90UH948-Q1
Other Parts Discussed in Thread: SN65DSI83

Tool/software:

Hi team，

Tier 1：Harman

1、What’s the meaning for tppos and tccs of DS90UH948 datasheet? Which part they are in timing diagram?

2、What’s the intra pair and inter pair skew guideline for openLDI/LVDS? I find

  • "High-Speed Layout Guidelines for Reducing EMI for LVDS SerDes Designs.pdf": 5mm for inter pair skew and intra pair skew
  • "slla038b.pdf", seems 1cm for inter pair skew and 5mm intra pair skew
  • So which one is right?

3、how long the routing length can be supported by DS90UH948 and SN65DSI83? Is 1m or 2m possible?

  • From "slla038b.pdf",it mentioned 5cm for 4mA output buffer.

4、Harman am comparing the driving capability between DS90UH948_openLDI_driver and SN65DSI83_LVDS_driver, so perform the simulation for the topology:

DS90UH948 or SN65DSI83 driver ->  ~6cm PCB trace ->DLPC230 openLDI receiver, as below simulation deck.

From the waveform at receiver, the voltage driven by DS90UH948 is much smaller than the voltage driven by SN65DSI83, why?

As I also compare both driver voltage without PCB trace, and find the voltage of  DS90UH948 is much higher than the voltage of SN65DSI83, why the voltage is much smaller at receiver for DS90UH948?

  • 0
    TI__Mastermind 30665 points

    Hi Alan,

    I'll answer the simulation questions first. 

    Alan Xia said:
    From the waveform at receiver, the voltage driven by DS90UH948 is much smaller than the voltage driven by SN65DSI83, why?

    Without having Harman's simulation model, I do not have a valid explanation for this behavior. Can you email the PCB model Harman is using so I can perform an identical simulation on my machine? From a first look, I agree that this looks odd.

    Alan Xia said:
    As I also compare both driver voltage without PCB trace, and find the voltage of  DS90UH948 is much higher than the voltage of SN65DSI83,

    Yes, the driver strength of the UH948 is stronger than the DSI83. This indicates to me that the PCB model is not equivalent for the UH948 and the DSI83. 

    Best,

    Jack

  • 0 in reply to Jack Scherlag
    TI__Mastermind 30665 points

    Hi Alan,

    I've received the touchstone file from Harman. I expect to be finished with simulations soon.

    Alan Xia said:
    1、What’s the meaning for tppos and tccs of DS90UH948 datasheet? Which part they are in timing diagram?

    tccs is the typical skew between the differential pairs of the OLDI output for the 948

    tppos designates the position of the pulse positions of the bits. The typical value of <0.1UI means that the OLDI output for the 948 has the bits placed to <0.1UI. UI = 1/7 bit time.

    Alan Xia said:
    2、What’s the intra pair and inter pair skew guideline for openLDI/LVDS? I find

    I will post the OLDI guidelines below. Generally our team aims for length matching within 5mils for inter-pair and intra-pair skew.

    Alan Xia said:
    3、how long the routing length can be supported by DS90UH948 and SN65DSI83? Is 1m or 2m possible?

    The DS90UH948 and SN65DSI83 were designed to be on the same PCB as the TCON. A distance of 1m-2m between device and TCON has not been validated or recommended. If the customer can prove that it works with simulation/testing, then they can use it as they see fit.

    Best,

    Jack

  • 0 in reply to Jack Scherlag
    TI__Mastermind 30665 points

    Hi Alan,

    I extracted the s-parameters from the Harman PCB touchstone file

    There is little insertion loss for the frequency range that OLDI operates in. Below 3GHz there is under -2dB of loss which will have a small impact.

    From my simulation results, I see that the 948 OLDI signal is larger than the SN65DSI83 OLDI signal

    Best,

    Jack

  • 0 in reply to Jack Scherlag
    Prodigy 80 points

    Hi Jack,

    thanks a lot!

    It seems receiver DLPC230 model termination issue.

    • If  use 50ohm R to GND as receiver, same result as you.
    • If  use DLPC230 as receiver , the voltage at receiver is only ~155mV, as the waveform in last mail
      • Also sweep several series resistor value, still not work
      • So can you please help check this?

    one more question, as DS90UH948 and SN65DSI83 were designed to be on the same PCB as the TCON, which TI LVDS/OLDI chip is designed for long distance board to board transmission(1m~2m)?

    Thanks

    Wenjun

  • 0 in reply to Wenjun Tao
    TI__Mastermind 30665 points

    Hi Wenjun,

    OLDI is required to be terminated with a 100Ω differential impedance. Both the DS90UH948 and SN65DSI83 datasheets show the 100Ω termination in the application diagrams.

    Wenjun Tao said:
    one more question, as DS90UH948 and SN65DSI83 were designed to be on the same PCB as the TCON, which TI LVDS/OLDI chip is designed for long distance board to board transmission(1m~2m)?

    What's the application you are developing for? Is there a reason the DS90UH948 and the SN65DSI83 can't be on the same board as the DLPC230?

    Best,

    Jack