TMS320F2808: Similar Program but Different Behavior

Nikhil,

Would it be possible to measure the resistance to both VDDIO and VSS for the GPIOs you are using for the CANRX/TX in your system?  I'd like to make sure there is no short and that the IO is good before looking further in the design for an issue.

Since you are using Uniflash I assume the flash programmation passed correctly, perhaps you could use Uniflash to execute a verify only on your flash image, just to make sure that everything is correct on the bad devices.

Best,

Matthew 

Hello sir,

Thank you for your reply.

Is it making sense of checking resistance of GPIO with VDDIO ?

Because i have tried 2 micro in different boards in which micro 1 works fine with both boards and micro 2 doesn't.

(refer the image attached to get what i tried)

Definitely i will check and let you know the result of resistance of GPIO with VDDS and GPIO with VSS.

Is there any thing else to check?

Unfortunately we do not have firmware to debug. we have only .out file to download.

- Nikhil

Nikhil,

Yes, I want to see if there is physical damage to the pin in question that would show up as a short to ground or power.  It would be odd that this should occur on 15 new devices from TI, but want to rule it out first.

Regarding the .out file, I think you should be able to use this with Uniflash and instead of programming the device, execute a "verify" that will confirm that the image on the device matches the .out file.

All devices are tested before shipment to ensure they are good devices, so again, it would be unusual for any device, and certainly 15/20 to have a hard fail like this.

Best,

Matthew

Hello Sir,

I have checked the VI characteristic in VI scope. 

I have selected CANRxA and CANTxA ref with VDDIO (pin no 3) and VSS(pin no 2) .

Please find the VI images from the below link.

https://drive.google.com/drive/folders/19kzm58OwQHqs1_kv3EzMu-qaKMRaHQA8?usp=sharing

if your not able to access the file please mention your mail i will forward to you.

Nikhil,
I'm going to send you a friend request through the E2E forum tools.  This will allow you to share/attach files privately.

Best,

Matthew

Nikhil,

   Please elaborate exactly what error you are seeing. Are you not seeing any activity on CANTX pin? Have you looked at the debug tips in www.ti.com/lit/spra876? 

Would you be willing to share the schematics of your board privately with me? If you don't want to share the whole schematics, please share just the CAN portion. I want to see the connections from the CANTX/CANRX pins all the way to the bus.

I looked at the V-I curves you had sent Matt. The I/O buffers on the pin are sourced by the VDDIO pin, not the VDD pin. 

I agree with Matt. 15 out of 20 devices failing is very unusual. 

Hello Sir,

We are servicing custom made hardware board here. One of our customer's board facing the following issue after sorting out the hardware troubles.

To get find out the trouble we have taken 2 BOARDs. Board 1 and Board 2. Then we loaded or programmed both board controller with same file (file.out) .

When we powering the board BOARD 1 gets running as normal and BOARD 2 showing error. The error is nothing but LED indication with relay. We exactly don't know why this error is happening and which section it throwing error because we do not have the firmware to debug the issue.

Then we did some experiments on board by interchanging BOARD 1 micro1 with BOARD 2 micro2. After replacing we noticed that the failure micro2 from board 2  in which we got the error indication is still showing the error indication in Board 1. 

I mentioned CAN because if we are getting error indication on board the CAN communication will not work/ get disabled. But in the same hardware if we are replacing Micro 2 with Micro 1 , it will work without an error indication. 

The connection between the I/O with can IC is a direct connection without any resistor. Please find the attached diagram for details.

sharing the schematic i have to get the permission from top level. Regarding this i will let you know the details.

Note : the board also deals with other peripherals such as adc , serial communication etc.

Thanks ,

Nikhil

Nikhil,

   Basically what you are saying is that the error follows the MCU. And it appears the error is "permanent" in the sense once a specific device exhibits the failure, it never works again even if you solder it on a known-good board. Is this correct?

Did the "bad" devices ever work properly? Or have they shown this issue right form the first time they were put on a board?

Nikhil,

    It might be more time efficient to discuss this over a call. I am currently in India and will be able to support a day-time call. Please send a friendship request and I will setup a call.

Hello Sir,

Q : what you are saying is that the error follows the MCU. And it appears the error is "permanent" in the sense once a specific device exhibits the failure, it never works again even if you solder it on a known-good board. Is this correct?

A : Yes sir. Once the MCU met the error condition it wont come out even we replace the hardware with good board.

Q ; Did the "bad" devices ever work properly? Or have they shown this issue right form the first time they were put on a board?

A : We have purchased the MCU from mouser.com and the MCU is programmed first time on the board.

hi,

I have send you friendship request..  Please check ..

Nikhil,

   As mentioned in the call, you cannot have a series resistor on any power-supply pin. That is a big no-no.

Hello Sir,

Thank you for guidance.

As you mentioned in the call we are tracking all Power supply pin's connection with regulators. And we definitely remove resistors if there any and will replace with 0 Ohm resistor. I hope it will be fine.

Also you were telling to check with code compressor studio with our hardware.

Would you don't mind can you guide what to check? so that i can check and reply you with result.

We have code composer Studio 8.1.0.

we are using Uniflash with XDS100v3 to download the image.

Nikhil,

       What did you observe after shorting the series resistor on the VDDIO pin? This is very important for me to know. I still don't have a good understanding on exactly what the "error" is. Until we understand the nature of the error, suggesting future experiments is hard. The fact that you are able to program the "defective" devices tells me that the device is largely OK. 

Nikhil,

   Any update?

Sir,

We have checked as per your suggestion by removing resistor from vdd pins. But the result is same as previous. The device which showing error indication is still showing the error.

For your better understating i will be sending videos of our OK and Error Board power on status in your personal message box.

Please check and let me know if your still confused.

My suggestion was not to "remove" the resistor from the VDD pins. My suggestion was to short the resistor on VDDIO pins.  I did look at the videos you had sent, but I wasn't sure what I am looking at.

Sir,

We replaced 100k resistor with 0ohm resistor. Sorry for misunderstanding.

Nikhil,

Is it possible to probe pins 6/7 on the F2808 on your earlier schematic both with and without the CAN transceiver populated on the board?  I'd like to see if we see any irregular voltage levels (not 0 or 3.3V) that would imply we aren't getting the correct values on the pins.

Another thing to try would be without the CAN transceiver populated(or at least the connection to the C2000 broken) if we see any difference in the good/bad C2000 on those pins.

Would you be able to share the PN of the transceiver as well?

Best,

Matthew

Nikhil,

      Please share the complete schematics with me privately. In the interest of time, please initiate a return of 2 or 3 suspect units to TI.

Helo MathewPate Sir,

Definitely i will check the voltage levels of pin 6/7 on Device and let you know.

Pardon me sir am not aware of PN of the transceiver as well.

Nikhil.

Hello Hareesh Sir,

For sharing the schematic i need the permission from top level. I will enquire and let you know about it.

As per our previous telephonic discussion we are planning to return the DSP controllers to TI through mouser.

I will update you the status this.

_nikhil

Nikhil,

    I will await both the schematics and the suspected units.