Duris_I

Hello,

I see probably some problem also with that USB adapter. Do you have another one for test? Simply I don't understand if you have good results with higher speed than lower speed, with this lower you have problems.

 Hello,

I'm sure there was no problem with USB adapter.

with the same program just by changing baud and trigger i'm able to get results for lower speed (9600, 19200).

with 38400, only read.
with 57600, nothing.
with 115200 , able to read and write, but errors are coming frequently. Some times read and write happening in a cycle, sometimes not.

With Regards,

Chaya Naga Sumanth

Kaulquappe

Hello Chaya,
the higher the baud rate is, the more important are the terminating resistors. I recommend to insert them.

Regards,
Kaulquappe

 Hi,

I will consider your suggestion.

Please see the attachments.

I'm using simulator for slave.
Pic's shows the logging of data received and transferred from master.
You can see RX and TX with respect to slave.

Picture with name "1"
you can easily observe that write operation is not happening properly, where as read is happening perfectly.

Picture with name "2"
You can observe both write and read operations are executed properly.

Right now, i don't have any problem with read operation at 115200 baud.

I have converted write masters to read masters and then executed the program, there was perfect transfer of data, 16 masters are able to read one by one, All I changed in the master block is MODE only (making sure that data length is valid).

Right now i'm not using any terminating resistors, still read operation is executed properly, all the problems are with write operation.

Only i thing i don't understand is , at 115200 baud , read is happening properly, write is not happening properly.

At low baud rates (9600, 19200) read and write are happening perfectly in every cycle.

help me solve this.

With Regards,
Chaya Naga Sumanth

Hello,

you can also do simple other thing (also as I already sugested to you change USB adapter), use another Slave SW simulator only for test with that speeds, but via same adapter. This is also what will be interesting. Simply you must do something exactly to find problem.

Kaulquappe

Hello Chaya,
due to your screen shots the slave did not recognize the request as no response from the slave is recorded. I think a character of the request is destroyed, so that the slave does either not recognize its slave address or the slave recognizes a bad CRC.

Once again: insert the terminating resistors and/or change the USB converter.
You can omit the resistors as long as everything works fine. If not, the first thing you should do are always the terminating resistors.

Destroyed characters are also influenced by the timing on the line. So it is possible that some baud rates work better than others.

Regards,
Kaulquappe

 Hello,

Please the attachment.

As mentioned in the manual of S7-1200 i have added terminating resistors.
Still my WRITE command is not working properly.

I have changed the converter too, still WRITE is not proper.

In both the conditions, READ command works fine.

Timing on the line is maintained for the minimum period to fetch or send data.

As you said some baud rates might work better than others, but my doubt is
"READ is happening at 115.2Kbps, WRITE is not  happening properly."

I have followed same method of switching of MB_MASTER for READ and WRITE.

I'm using 10Hz trigger, counting the pulses and resetting, comparing the count value to switch REQ of MB_MASTER.

I even decreased Response time to 100msec (transfer of 300 words between master and slave easily happens with in this time with 115.2kbps baud).

Help me sort out this problem.

With Regards,
Chaya Naga Sumanth

Hello,

again, in case that you've test also another adapter, did you try select another MB SW simulator? Also I recommend to you making tests with simple bit memory for request, not timer. In this case I see simply lot of not detected problems. Simply with bit memory request you can then after correct or error receive simply reset that request memory and see results. Also I'll be very carefuly decrease that timeout, simply with proper condition looking to DONE you will have good solution. You can also include some counters about requests and bad responses. If you'll have relevant results then you can make your code to next level.

Update: in my opinion I'll be run some other SW (it can be any kind of serial SW which can show incoming message, not only Modbus SW) and check that message coming from your PLC. With any of your need communication speed these results can be interesting for me.