two controller TM4C 123gh6pz are configured and used in UART , during functioning one controller gets hang .

Amit Ashara said:

Hence the further of the post for details.

Which could be severely reduced (perhaps eliminated) by the introduction of, "Proper Forum Posting Guidelines!"       (as several here - have long suggested - even submitted!)

Have you thought to "track" the number of times you've had to invest time/effort to, "Request basic details?"       Bet that's quite high - and growing - where's the "cost savings - or containment" via that "unguided poster, (NON-Method?)

How many neat projects - far better - deserve your time?        Sacrificed (instead) at the altar of, "Does not Work" or, "Controller gets hang!"

My past "Poster Guideline" fully detailed - what clients, investors & myself agreed to be - a "workable" solution.     And while that post quickly surpassed 500 (or so) "hits" - after its "30 days" of limelight - it drifted off into "forum oblivion."

While the "suggestion box" may (appear) to be open - any adoption/adaptation of "NIH" suggestion - is (and remains) unlikely to, "see" the light of day!    Pity that...

Hello Amit,

OK

Let me brief the product and details as

I am using these two controller  in Electric discharge machine  one is for HMI / Keyboard  and  2nd one is for main EDM process control in the process control it has

ADC input , Pulse generation , Servo loop control, I2c line  Input out put control , fast interrupt base input for servo loop control.

these two controller are connected through  UART and with differential communication.

The Noise level in the machine is very high.

The Machine is run for a one hour and then the communication between these two controller lost , but these two controller runs independently  as the keyboard works. and servo loop of main controller is running . The display update date is transferred from main controller to HMI controller , since the lost in communication the display will not update ,

 I think these details will get some idea , if any more details needed pl. ask.

Regards

Sanjeev

Still not enough details, so one can only guess.

But IMHO you have two problems. The first - your communication protocol seems not robust. You need to cope with corrupted and lost characters. So implement, for instance, frame checksums, acknowledge messages, timeouts and retransmissions. Second, you can try to improve the situation by "EMI-hardening" your communication, i.e. shielding, current loop, reduced baud rate or an EMI-proof medium. Industrial control units often use optical fiber lines in such cases.

Poster f.m. - as long is his habit - "Stole my thunder" w/his rather complete listing.     (thanks for that - f.m.)

Beyond his comprehensive listing - is it possible to add physical separation between the MCU board (nearest) the (offending) motor/actuator "action?"     Inverse square law often enables such "distance" to succeed - removing the requirement for the many defensive tactics poster f.m. suggested.

Another "free" technique - if there are allowable, "down periods" (those in which NO commo is required) you may wish to regularly/repeatedly, "Reset both UARTs - to insure their "readiness."    In some of our "high noise" applications (often those in which the "high noise" is new - due to unforeseen additions) we may add a "commo heartbeat" which allows recognition that the serial link is, "alive/well."    Should that "heartbeat" be lost - UART Reset is ordered.

Suspect that for (most) applications including (some) of f.m.'s suggestions (where budget allows) will yield best results.    Tricks offered by this reporter enable (sometimes) the "saving" of an "over-challenged" commo link...   (escaping cost/size additions)