hi
I use below isolator for connect pc to pc but dont work this circuit?

 

Because there are too many unknowns.

 

Because there are too many unknowns.

what is unknown?
Optocoupler is CNY17 .

 

What is J1 computer?
What is J2 instrument?
What software are you running?
What are the COM protocols on both J1 and J2?
What data is being transmitted from where to where?

 

What is J1 computer?
What is J2 instrument?

this picture is in below website
http://www.123microcontroller.com/Hardware-Interfacing/OptoCouple-with-RS232
perhaps J1 and J2 is a symbol !

What software are you running?

Hyper Terminal

What are the COM protocols on both J1 and J2?

RS232

What data is being transmitted from where to where?

this circuit don't work
Can you show me a circuit for RS232 isolator?

 

For starts, you have to use no hardware flow and/or handshaking control between ports for this device to work. Also you must make sure that your RTS signal on each of the two ports is logic high (a zero, in this case) sot that it can deliver up to +12V (+15V in some cases). And finally, whatever opto-isolator chip you're using must be sensitive enough for it to work with the meager 5ma that the RTS is normally able to deliver.
There might be other factors playing here, depending on your hardware, software, and application. But we'd need to know far more details than the ones you've just given.

 

Can you show me a circuit for RS232 isolator?

Just Google "RS232 optoisolator" and you'll get plenty of answers. Some self-powered, like the one you've just given. And others powered through an external power supply.
I'd use the latter option if I were you, since it would be far more robust and reliable.

 

As Cmartinez mentions, you need to set the hardware to Null Modem, you do this by jumpering the 1-4-6 & 7-8 pins at each end.
Is there a particular reason you are wanting isolation?
The signal for RS232 is a ± voltage signal.
I use a diagnostic tool such as RS232 HEX Com.
Max.

 

Is there a particular reason you are wanting isolation?

I had a nightmare with a ground loop two years ago while trying to interface a CNC ... that's usually the most common reason why people set up optoisolation in RS232 devices, other than special safety controls requested by the user, of course...

 

For starts, you have to use no hardware flow and/or handshaking control between ports for this device to work. Also you must make sure that your RTS signal on each of the two ports is logic high (a zero, in this case) sot that it can deliver up to +12V (+15V in some cases). And finally, whatever opto-isolator chip you're using must be sensitive enough for it to work with the meager 5ma that the RTS is normally able to deliver.
There might be other factors playing here, depending on your hardware, software, and application. But we'd need to know far more details than the ones you've just given.

but work properly when I connect directly two pc Together without OPTO_ISOLATOR (with 3 wire , Rx , Tx and Gnd) , so The problem is not the RTS , do I am in wrong?

 

I had a nightmare with a ground loop two years ago while trying to interface a CNC ... that's usually the most common reason why people set up optoisolation in RS232 devices, other than special safety controls requested by the user, of course...

I usually overcame this by obtaining the power for the PC from the CNC machine itself, I found most problems of blown communication IC's was when a different source of power was used, Ground to ground voltage difference etc.
Max.

 

Is there a particular reason you are wanting isolation?
Max.

I want to connect my pc to a instrument that have different power supply , of course After the success at this stage

 

but work properly when I connect directly two pc Together without OPTO_ISOLATOR (with 3 wire , Rx , Tx and Gnd) , so The problem is not the RTS , do I am in wrong?

I don't think you are getting the right ± (plus to minus voltage) signal transition?
Max.

 

I want to connect my pc to a instrument that have different power supply , of course After the success at this stage

I have done this for many years, as long as the commons of both are connected it should not matter.
Max.

 

Max nailed it.
RS-232 has to go -V to +V.

 

I have done this for many years, as long as the commons of both are connected it should not matter.
Max.

You are right , I before connect directly my pc to device and Could not create any damage but don't talk together !
this device have a Transducer that is rs232 isolator and device easily talk with pc
So I certainly should use of this type Transducer

 

I usually overcame this by obtaining the power for the PC from the CNC machine itself, I found most problems of blown communication IC's was when a different source of power was used, Ground to ground voltage difference etc.
Max.

In my case, the PC was connected to the same 110VAC power line as the servo driver I was trying to interface to. But it didn't matter if the grounds were connected close together or not... I'd get huge error readings unless one of the sides of the shield was left unconnected on the three conductor shielded cable that I used between the two DB9 connectors. So after connecting pins 2, 3 and 5, and leaving the shield connected only on the PC side of the cable things became acceptable, though not perfect. I'd still get one transmission error every 100 characters or so, which I had to manage by introducing CRC and communications monitoring techniques.