100 KHz Signal Transmission Over Long Distance

Rather than voltage turn it into either a current signal or a fibre-optic signal.

Another way is to run it in a balanced and shielded feeder cable.

Any other ideas?

Additional to PWS post. You could use a radio coaxial cable (or ethernet "triax" cable), or driver and receiver to RS422 standard, which are differential transmitter and receiver for use with 2 core 120 ohm approx impedance screened cable. RS422 or similar RS485 (bi-directional) are commonly used for industrial PLC remote communication. If your "5V digital" means a plain TTL IC output, you must accept that you need interface drivers/receivers for such a distance.

For this distance I also think about the good old unshielded twisted pair configuration. Of course you need to convert the signal to differencial, if it's not already, for proper common mode noise rejection. S.M.

Please tell us more about the nature of the data, and the cost sensitivity.

Do you just need to detect "Signal On/Off"? Are you counting pulses? Is it carrying nibbles/bytes/words of encoded data? Is it bi-directional? Is there any error correction protocol?

You could go for top-of-the-range (which in this case I'd say would be a fibre optic connection, but it would obviously need converters at either end), but if less data integrity is needed, something cheaper would probably do the job.

Hi

thanks for all reply.

Actually, my requirement is as follow:

I need to transmit 100Khz TTL pulse as count pulses for PLC's 24 volt input.

Any suggestion will be appreciated.

Thanks n Regards.

A PLC running at >100kHz? Good grief!

Presumably it's a high-speed counter input (e.g. the Allen-Bradley 1769-HSC, which can count at up to 1MHz).

Can you give us the PLC input characteristics, or (better) a link to the manufacturer's datasheet/user manual?

It is always best to write what you want at step 1 !! Wanting a 24V signal from a 5V one is a different situation!

Obviously, you need a "level converter", so we are not doing "cheap and cheerful" - the first good practice is to convert the 5V to 24V at the signal source and send the stronger signal through the "noisy" cable path.

Some information about the 24V "receiver" would help (from the PLC data sheets):-

1. Does it have a 24 VDC power point? Or is the input tied to +24V through the input via a resistor? If it does, it may be possible to avoid a separate power source to get 24V.
2. What is the input load resistance presented by the 24V input?
3. What are the switch voltages at the input? Usually stated " 0 state (off) < 5V : 1 state (on) > 15V" or similar?
4. Do you know what your pulse length is? 100 kHz may mean "more than 1/100000 second = 10 microseconds between pulses" NOT "100000 pulses in each second". The second could mean " 10000 pulses in 100 nanoseconds, followed by 1/10 second silence, repeated 10 times in a second".
5. A pulse length of 5 microseconds seems good - remember, if each pulse is only 1 microsecond long, it has less energy and is less able to penetrate a noisy environment. In addition 1 microsecond will need more bandwidth for distortionless transmission, causing more vulnerability to environmental noise

Hi,

PLC input is pulled up with +24V via a resistor internally.

I have to supply 100Khz pulse at 50% duty cycle.

As at source point a level converter is required from 5v to 24v, I am thinking to use a optocoupler.

Which optocoupler ic would do the job at 100Khz?

Further comments are welcome.

Regards.

1. So how much current is needed to pull 24V input down to 0 V?
2. And as written previously, what does the PLC input data sheet say about the "0" and "1" input voltage limits of the 24V input??
3. What is the capacitance of your intended cable?
   

Use a twisted wire pair inside a shield. Ground the shield. It may not be EMI proof, but will probably do the job. Some of the better shields have two braids.