This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TMDSHVMTRPFCKIT 28035 Piccolo-ControlCard connection problem

Other Parts Discussed in Thread: TMS320F28035

hello,

I've got some problems with the following kit:

TMS320C2000
TMDSHVMTRPFCKIT 28035 Piccolo-ControlCard

I went step by step through the "How to run" and "HW" guide.

My settings:

J3,J4,J5 [all main] are populated J5[M3] is not.
J2[main] is connected between the "Bridge" and the middle pin.
BS1 and BS5 are also connected via banana-plug.
(a) J9 is not populated due to the fact, that the DIMM-Card(28035) is in J1.
This confuses me, because the "HW" guide says, that:

'[Main] J9
JTAG TRSTn disconnect jumper, populating the jumper enables JTAG
connection to the microcontroller. The jumpers needs to be unpopulated
when no JTAG connection is required such as when booting from FLASH.'

But this would be the complete opposite of (a) and I have to poulate J9 for programming the controller on the Piccolo ControlCard.
despite that I need the xds100 Emulator and therefore the JTAG all the time to work with the kit via the pc.
Is that right (referring to the J9-topic)?

The cable for the USB JTAG (XDS-100)Emulator is plugged in the usb port on the pc and the kit. (LD1 is on)
I connected the DC power entry [M6] (and also tried without it, thinking that the supply from the pc-usb should be enough.)
With the DC power, I think a pre-flashed program is instantly running and turns the fan for cooling the inverter on maximum speed.
In order to stop it and change running programs I have to connect the kit.

My settings for the target configuartion in CCS:

'Texas Instruments XDS100v1 USB Emulator

TMS320F28035'

CCS version: 4.2.3.00004


There seems to be a connection problem. I receive the following statements during debugging (it is one of the ti sample
programs 'HVMotorCtrl+PfcKit_v2.0'):

marginal note:

After importing existing project from the controlsuit I get an Error: (see below)



(For half the projects)

I couldn't rebuild the other successful imported half of the projects. It only works after installing a newer code generator version, but I have to do it manually.
(The update function in CCS creates only an empty folder.)
I still don't really know what exactly to do to get all the examples working.

My CGT version is now:TI v6.0.1.

After starting debug, the console opens with 'Error 1135' (see below)



Then I reseted the emulator and debug it again. This time 'Error 1041' shows up.



The device manager shows:

-XDS-560 JTAG Emulator (with warning sign, but it seems to be irrelevant)
-a new COM Port appears
-TI XDS100 Channel A
-TI XDS100 Channel B

I've checked the eeprom programming; the values are:
VID:0403
PID:A6D0

Everything seems to be ok, concerning the hardware connections.

What power supply is exactly necessary to run and program the Controller on the Card?
(At the moment I don't want to connect the kit to a motor, so it's all about programming.)

Are my mentioned hardware-settings ok?
Or is there a complete other problem I'm missing?

best regards and thanks





  • hello,

    Me and a fellow student  figured out, how to solve it.

    You have to set the full access option in the folder properties of ccs and controlsuit even when you're admin and not just a simple user. (This also works for the DRV8301-HC-C2-kit.)

    J9 have to be populated otherwise the debug woud not work. (by the way "the how to run" guide says:'If ISO-DIMM card is used, then do not populate [Main]-J9. ' on page 4)

    best regards

  • Well, it's still not working for me--i still have the connect error -1041 and I have try again and again to make a connection to the F28027. What is particularly maddening is the fact the this board and the one before it were once working just fine with the CCS tools. I originally thought the early board somehow got damaged, but now it has happened twice.

    The JTAG checking tools reports all is well with the connection. The FTDI tool, FT_Prog is able to read and program the flash associated with 2232D USB controller I am using on the board.

    I pretty much convince that the problem lies within the CCS tool--the flaky, unreliable CCS tool we are all saddled with.

    Has anyone gotten a straight answer from TI as to why problems like this occur? Forum entries extend back several YEARS complaining about connect problems.

    DE

  • Hello Team,

    Is there any insight that you can offer to Mr Ehrenberg regarding his error -1041 issue?

    Regards,

  • Few things to check,

    Make sure only one emulator is connected on the machine,

    What exact version of the board are you using,

    Software version v2.0 is targeted for the yellow PCB board, Rev 1.1  (this is the latest version of the board)_and

    Software version v1.7 is targeted for green PCB Rev 5, 

    J9 need to be populated, if it is not populated JTAG connection will not work,

    Can you identify the board version, Control card version, your machine OS?

    Are you able to run the QSG GUI?

  • Manish, the PWB is my own design--shamelessly copied from TI's schematic of the Piccolo docking station. There is only the one USB connection through the dual channel FTDI 2232D device. Even though CCS refuses to connect, the second channel, configured as a serial port contrinues to function properly--I can command the previously programmed (in flash) functions and the hardware responses appropriately. The XDS100 v1 channel A and channel B appear in the Windows 7 device manager along with the additional (and still functioning) COMxx port.

    I tried establishing a download link using CodeSkin's C2Prog for a CCS generated Intel hex file--no joy, same error message, -1041. We even tried a serial download (as apposed to JTAG) of the image, but that feature is apparently not support (beit software or hardware).

    I have yet to try the QSG interface. However, I dusted off my old docking station and Piccolo F28027 SIM stick and that combination behave normally. It appears that the F28027 uCs have degraded after a dozen programming cycles (not the 100k claimed in the literature) and I need to replace the uCs. If I had an exact timing spec for the JTAG connection, I could check the timing on the four JTAG leads to get some insight into the problem. As it stands, my only comparision is to the still functioning docking station/stick combination.

    DE

     

  • David,

    I am a bit confused, the tile of the thread says HVMTRPFCKIT with F28035 control card. From your recent post i understand you have a docking station? which is a spin from the docking station that TI has. Did you modify the schematic or gerbers when you did this.

    Also you are saying the control card on the new docking station is replying to the serial commands, that means the serial channel is working but the JTAG will not connect.

    There is a lot of  difference between docking station and the kit wrt to emulator, the one on HVMTRKIT has isolation for the USB,

    Regards

    Manish Bhardwaj

  • Manish, I'm using the same circuit as that found on the Ti docking station on my custom PWB. The uC is the F28027. The "damaged" uC would continue to respond on the serial channel (of the two channel FTDI 2232D), but the JTAG connection though CCS5 could not be established and I was unable to flash program newer code. Apparently the production problems of Piccolo uCs have been fixed and I was able to get some more units. I replaced the "damaged" unit with a fresh, out-of-the-box device and can again establish communication via CCS5--and load new code. Why the old one fail we may never know . . .

    But even with the fresh uC, I still see messages appearing on the CCS5 console flagging communication problems with the device. I largely ignore these now since I can conduct debugging through the serial link.

    DE

  • Manish, yet another uC has failed--what was once working is now exhibiting error -1041. The failure occurred shortly after a firmware upgrade of the flash code. I don't know whether this is significant or not, but the last flash operation before failure was the erase and programming of sector C of the flash. I'm using this sector to record user modifiable parameters. As before the console interface continues to work.

    What is error -1041? What conditions cause its invocation?

    DE

  • Working setup stops working after flashing point em to think if you programmed the flash keys accidentally ?

  • Also when you open the ccxml file, you can choose to Test Connection, this will open a window that will display some more information on the JTAG