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.

Control LCD at 1-2 meter distance from the controller with parallel communication

Hi Team,

Can you please help us with our customer's inquiry below?

I would like to manage an LCD display (128 x 64 Pixels module) and to contain the costs of my application I would like to choose one with parallel communication. The LCD screen should be about one meter from the PCB that controls it and I would like to ask you, since I do not find documentation on ti.com, if there are techniques to manage parallel communication protecting it from any disturbances.

This is the configuration
microcontroller >> digital isolator >> buffer >> 1 or 2 meters of cable >> expander port >> LCD
Is a buffer sufficient in this configuration, even in a noisy environment full of synchronous / asynchronous motors? the I2C so extended is not severely affected? Would digital isolators help prevent ground loops (always taking the capacitance of them into account) or would they be better than TVS diodes (maybe they have lower capacitance, I have to check) and use various ground paths in PCB layers?

Regards,

Danilo

  • Hi Danilo,

    Based on the specs provided for the LCD, I recommend forwarding this question to the digital isolation team. For displays with PCLKs of at least 5MHz, the FPD-Link Portfolio can help with mitigating EMI and reducing video control signals going from board to board.

    Regards,

    Jack

  • Hi Jack,

    Our customer would like to directly connect the I/Os of the microcontroller to the parallel input of the LCD using parallel cable (1 to 2 meter) to reduce cost. He would like to know if using a buffer sufficient in this configuration, even in a noisy environment full of synchronous / asynchronous motors? 

    Regards,

    Danilo

  • Motors are very noisy, so you are likely to need differential transmission, i.e., LVDS. It would be possible to use many LVDS transmitters and receivers, but  FPD-Link makes this simpler by (de)serializing many parallel signals through a single LVDS connection.

  • Hi Danilo, Jack,

    What's the impedance of your 2 meter transmission line? If you want any chance at a buffer working, you'll need to properly terminate the transmitter to match to the transmission line. There's some information regarding this in our FAQ here: [FAQ] What happens when I connect a logic device's output to a 50 ohm transmission line?

    In a noisy environment, I would recommend something differential like LVDS to eliminate common mode noise in your digital transmissions.  You might be able to get away with a standard logic buffer, but being single ended it won't be able to filter out CM noise. Using a Schmitt-trigger input device will help some with noise, but there's a point where no simple buffer will be able to handle it.

    It seems to me that saving a few cents here wouldn't be worth having to rebuild the entire system if you find out a buffer isn't good enough to handle your application's noise levels. If you can setup a realistic test and measure the noise levels, it would help in deciding what can or can't be used.

  • Hi Emrys and Clemens,

    Thank you for your valuable feedback. Please see the comment of our customer.

    I know how expensive it is to produce a PCB, especially if it has many layers. slightly increasing the cost of the PCD would lead me to prefer the use of an additional microcontroller (such as TMS320F28002x) connected to the CAN network, i.e .:
    Main board >> CANbus >> digital isolator + microcontroller + LCD
    since my application is developed on a CAN network (since this type of port is suitable for any environment and allows addressing). Using buffer and ser / des does not convince me, because I do not know exactly how to make a realistic test to measure the noise level, nor how to theoretically estimate it (if there are articles about it on ti.com, I would very much like to read them); moreover it would lead to complicate the Main board, occupy more PCB surface, ...
    In your opinion, is this the best alternative? or the only feasible solution?

    the article you posted is really interesting, it enlightened me about aspects that I considered "automatic" and gave me new questions:
    -in what situations is the Parallel Line cable still used today? (I have a 3d printer at home, with an arduino processor that manages an LCD with Parallel Line of just under 1 m)
    - you told me that not knowing the noise level, it is not possible to recommend an appropriate buffer. Going under 20cm of cable and redesigning my application, would I encounter any noise issues? (The LCD screen has a very low resolution 128 x 64 Pixels. I don't care if some pixels are not lit in the correct way as I think I refresh the screen 4 or 5 times per second or if an "event" is activated for the pressing a key, manageable by an interrupt)

    Regards,

    Danilo

  • Hey Danilo,

    -in what situations is the Parallel Line cable still used today? (I have a 3d printer at home, with an arduino processor that manages an LCD with Parallel Line of just under 1 m)

    I don't see parallel lines as much as I used to, but I still see them sometimes in board to board connectors and low-speed connections. Up until SATA became the standard, IDE hard drives were using parallel cables for ~ 133MHz communication. I wouldn't be surprised to see many of those still in use.

    I think most system designers use them to carry multiple serial signals instead of actual parallel data, so there's less issue with timing / cable length.

    Going under 20cm of cable and redesigning my application, would I encounter any noise issues?

    I'm afraid there's no way for me to know your noise levels, frequencies, or how much they will impact your system. This is something that would have to be tested experimentally in the conditions you plan to use the device.
    Transmission line effects for typical logic gates start around 12cm of t-line, so 20cm would have some reflections if the correct termination is not used.