RS 422 Not Talking

Second page

The 0.5 volt peak to peak signal is too low to meet RS 422 standards. This maybe because the driver is bad or is not really RS 422 compatible, the receiver is in correctly terminated with a differential impedance of 100 ohms or the wiring is incorrect. I have done and seen all these.

If it is a differential signal, the minimum sensitivity signal amplitude is about 300mV (general data for LVDS). For non differential it is, indeed, too low.

The server is undoubtedly using RS-232. Get yourself an inexpensive RS-232 to RS-422 converter. Past that you might be having an issue with the protocol used.

Is this a problem that has recently occurred from a previously working state or is it a problem you are experiencing whilst trying the set the system up?

What happens if you disconnect the connections from the server to the radio? Does the amplitude go up? This may tell you if there is a hardware fault on the server transmit side or the radio receiver end.

Can you post a waveform that shows the 5 bytes with the start and stop states as well.

Regards.

Are you displaying a differential signal: i.e. using two probes and displaying trace A minus trace B.

OR are you just showing us one leg (the other leg being the opposite).

OR is it the output of the RS422 receiver.

Are you sure the first screen is not captured using a divide by ten (usually bizarrely called a X10) probe.

Synchronous (with a clock signal) OR asynchronous (some thing like "9600 baud 7 bits one stop bit parity" type) comms.

Half duplex or (both signals running in opposite directions on the same wires) or full duplex (separate sets of wires for Tx and Rx) ?

I wish I knew how to get CR4 to accept larger higher resolution pictures. But tell us what the scope settings are vertical and horizontal (I think I can make out the 200mV per division and 1V per division vertical ones).

EDIT----- I've looked at the pictures in more detail

I think I can make out 1ms per division for the first screen and 2 ms per division for the second.

On that basis it looks as though the first screen shows:-

08 C2 08 02 HEX at 4.8 K Baud 8 bits no parity 1 stop bit

red = stop bit green = start bit . I don't think any other patterns of bits, parity, stop bits fit, but I could be wrong.

the second screen seems to show

55 HEX at 2.4 K Baud

Normally the Tx and Rx should both be the same speed.

Neither of the sets of data seem to agree with the values you have quoted, but check the comms. set up first.

Oops! I was so engrossed with the larger trace, I didn't see the wider one on the top.

Using your labelling system, I can see that with the white bits you can get 01 34 01 04. The fifth bit at the end is incomplete on the trace and is probably the 00. Got this by reading LSB to MSB from left to right, not the usual way around! I suppose it depends on how the OP got his data.

Regards.

I think you may be right.

Or even 4? 10 43 10 40 ?0

OR 4? 10 43 10 40 ?8

Which is getting close to what they should be (but backwards).

That would make the other data AA instead of 55 I think.

Are you sure the first screen is not captured using a divide by ten (usually bizarrely called a X10) probe.

One GA from me wouldn't be as rewarding as a good deed in the Electronic Engineer's paradise. How in the world, after so maaaany years of engineering, I haven't noticed (not that I didn't know what each mean) the discrepancy between X10 and ÷ ? I, probably considered (in my EE subconstient) that it is x10 attenuation...I don't know. Shame on me!

You are correct!

Clocking mismatch issues with the new model PCI card. Had to re-configure the data rate and clock-multiplier. Working now.

Another problem solved!

Tim Hawley Master Mech.

Helo Szwasta,

Your problem is simple.

You can only use one RS422 connection on one com port at a time.

What you need to do is add an addition RS422 comm port by using you PCMCIA port or add an additional serial I/O interface board to your hardware profile. This will allow you to use two (2) serial ports at one time.

Best Regards, Tim