Dear ola84,

after checking on site, reading process works on a right way but only when running only step1 ( until Network 4 )

There is a limit from the slave side, of 48 words for reading during each REQuest.

I have tottaly to read 170 Words, so i must read 170 Words in 4 steps.

When step2 is running ( Network 5 and 6 ), "done" bit is triggered but error 8200 occurs random.

It seems that a new REQuest is on, but in the same time, the previous request has not been completed yet.

Do you think a time delay in Network 5 is the right solution ?

2. Step 1 ( until network 4 ) works only if start adress is 40002 and not 40001.

Attached the code in order to help somebody else, maybe for reading at once and not in multiple steps

Kaulquappe

Hello Kosmous,
are you sure about error 8020 or was it a typing error?

Regards,
Kaulquappe

Dear kaulquappe,

you are right, the random error that occurs is 8200 as shown inside pdf code in Network 2.

8200 means : Port is busy processing a send request.

So i think, that a new REQuest command is given and in the same time, the old request has not been completed yet.

Do you suggest a time delay before a new request is given in my code ?

kosmous

Hello,

after checking on site, reading process works on a right way but only when running only step1 ( until Network 4 )

There is a limit from the slave side, of 48 words for reading during each REQuest.

I have tottaly to read 170 Words, so i must read 170 Words in 4 steps.

When step2 is running ( Network 5 and 6 ), "done" bit is triggered but error 8200 occurs random.

It seems that a new REQuest is on, but in the same time, the previous request has not been completed yet.

Do you think a time delay in Network 5 is the right solution ?

2. Step 1 ( until network 4 ) works only if start adress is 40002 and not 40001.

Attached the code in order to help somebody else, maybe for reading at once and not in multiple steps

Could somebody check the attached code in order to suggest the best solution for the above problem ?

Hi kosmous.

I see that you set the REQ param. on mb_master two times, but it's never reset. As I suggested earlier, you should reset that param right away after mb_master.

Else I think you got it, you just need to extend your state machine in to 4 steps. When you are debugging on the site you should watch the step-variable and see witch steps are executed.

Regarding the 40002, I don't know the reason for that. 40001 is the first holding register in modbus.
( which is function code 3 and offset 0 )

Page 2 bottom: No, you should not put a "done-contact" there, that would make your state machine halt at some point in the future, when a random communication error occurs.

Hello ola84,

thank you for your interest,

you ere right, but i already use the RESET to REQ immediately after Master call. ( See attached )

Just it was not appeared in the pdf file that i had attached.

The problem still remains. The communication is established, leds are blinking, the slave responds that communication is ok, but every 2 sec, error 8200 is appeared for 1 sec and dissapered again cyclically.

When i use "done" contact in Network 6 in series with " busy normal close ", 8200 error is never appeared, but do i lose data in that way ?

When only step 1 is called, everyhting works perfect !

May be i need to check somewhere the "done" bit before start step 2 ?

Ahh.... I see the problem.

Network 3: Your program insert the query parameters. and is jumping in to state 11.
Network 4: Since mb_master has not executed yet, mb_master is still not busy, and your state machine will continue to step 2.
Network 5: Your program will carry on and go to step 22.
Network 6: Same story as in network 4, and the state machine will go back to state 1.

I haven't look at what will happen in the second execution cycle, but there seams to be some black magic in the code that you want to avoid.

My solution to the problem:

- Swap place of network 3 and 4.
- Swap place of network 5 and 6.

Hello ola84,

many thanks for your suggestion.

You are right !

The only strange issue now is that when start adress to read the slave is 40001, error 8383 still occurs even if i try to read only 2 words. Error 8383 means :

" Error in the data address or address outside the valid range of DATA_PTR "

When starting adress for reading process is 40002, no error occurs !

Attached is the structure of the adress that slave demands.

Do you see anything strange issue about 1st reading register 40001 ?

-------------------------------------------------------------------------------------------------

By the way, to help somebody else, my communication process worked with MODBUS blocks of TIA PORTAL and not with MODBUS ( RTU ) blocks.