• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 19 subscribers
  • Views 1501 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.

Long distance transmission of CMOS single end output

Satoshi
Guru 19495 points
Other Parts Discussed in Thread: SN74AHC14

By conditions, can CMOS single end transmit long distance pattern?

Customer want to use SN74AHC14 for 1m pattern and difficult to change differential line for customer reason.

Although Below application note is described that 1m CMOS line is not recommended by Chapter 1,

if there way about condition for possible to use 1m CMOS line, please let me know.

※If CMOS output is very low speed(6.7kHz), is this condition become to use long distance?

Best regards,

Satoshi

  • 0
    Guru 243980 points
    There are people who try to use long lines driven by general-purpose logic devices. In the lab, it often works; in the real world, or even in an industrial environment, if often fails, especially when using higher frequencies.

    6.7 kHz is quite slow, so it should be possible to add a heavy low-pass filter to filter out most noise (try an RC filter at the receiver). This will also slow down your signal edges, which reduces reflections.

    But noise isn't the only problem; how do you handle ground level differences?
  • 0 in reply to Clemens Ladisch
    Guru 19495 points
    Clemens Ladisch-san

    Thank you for reply.

    I will confirm customer's GND level.
    Sorry for mistake, correct line distance is 5m.
    Even if 6.7kHz, Is 5m line not realistic for single end?

    Best regards,
    Satoshi
  • 0 in reply to Satoshi
    Guru 243980 points
    5m is worse than 1m.

    You can look at the received waveform with an oscilloscope (and it will probably be fine), but this ignores any additional noise present at the place where the devices will finally be installed. If this isn't some industrial environment with much noise, and/or you have proper shielding, it might actually work. But nobody here can guarantee anything.
  • 0 in reply to Satoshi
    TI__Guru* 96591 points

    Satoshi-san,

    Something to consider here is that a 6.7 kHz logic signal is not as simple as a 6.7 kHz analog sine wave.

    In the analog case, you can determine delay and loss through the transmission line for that particular frequency only - and at 5m, it wouldn't be a problem at all.  A 6.7 kHz sine wave has wavelength of ~44km, which is so much longer than the 5m cable that it shouldn't make any difference, and the loss/delay at this range would be negligible.

    In the case of a logic signal, however, the analog frequency components are significantly more complex.  I don't want to get too technical here, but a square wave is basically the sum of many analog sine waves (please see this site for a graphical explanation).  Each sine wave has a different frequency, with the primary/largest amplitude one being the square wave frequency (in this case 6.7kHz) and the other components being higher frequency.  These higher frequency components can create problems even in relatively short cables such as you have described.

    If your logic device switches from 0 to 5V in 1ns (which is not uncommon at all), the bandwidth required to transmit your signal (and keep its original shape) now increases from 6.7kHz to ~350MHz (using a rule of thumb that BW = 0.35/(edge rise time) ).  At 350 MHz, the wavelength is ~0.86m -- and now your 5m cable is causing problems.  The high frequency components of the signal and the low frequency components of the signal will reach the distant end at different times (known as 'group delay' for logic systems), which will create some issues by itself.  Since the cable is likely not terminated at its characteristic impedance, you will also get significant reflections that could end up damaging the transmitting logic device.

    Short answer is - Clemens is right. A well shielded 5m line in a quiet (electrically) environment will likely work fine, but that's no guarantee that it will work always and anywhere.