Problems with JTAG

[anm]

Posted on May 14, 2009 at 09:12

Problems with JTAG

[jj]

Posted on May 17, 2011 at 13:10

Need more info to best assist:

a) JLink sounds like IAR/Segger - if new that should work. However to be sure have you another board with which to confirm the JLink is ok? b) Is it possible that you have ''re-purposed'' the JTAG pins prior to using your JLink? (likely by some ''non-Jtag'' means) c) Always advise that you build ''several'' proto boards - enables you to confirm that problem persists across multiple boards as opposed to a bad connection on a single/first board. Recheck continuity between every JTAG pin - especially @ the micro pins themselves.

[anm]

Posted on May 17, 2011 at 13:10

Hi!

I made a pcb with a STM32F103CB and JTAG Pins. Now I tried to program it via JTAG, but I get the following error: Could not find supported CPU Core on JTAG Chain Bad JTAG Communication: Wirte to IR: Expected 0x1 got 0x7 (TAP Command : 2) @Off 0x5. The JLINK Device says, that the operating voltage is OK and the power consumption is ok. What could that be? I also checked the wiring on my board (VDD,VCC,JTAG-links are correct). I don't know what to do next :( Greets from Austria Andi

[anm]

Posted on May 17, 2011 at 13:10

Thanks for your answer.

Quote:

On 01-05-2009 at 16:14, Anonymous wrote: Need more info to best assist: a) JLink sounds like IAR/Segger - if new that should work. However to be sure have you another board with which to confirm the JLink is ok? Its an IAR Jlink (the yellow one). I have a development board, so the JTAG is working properly.

Quote:

b) Is it possible that you have ''re-purposed'' the JTAG pins prior to using your JLink? (likely by some ''non-Jtag'' means) I haven't changed the configuration of the controller now. I just soldered my PCB, plugged the JTAG in and tried to program it - unsuccessfully...

Quote:

c) Always advise that you build ''several'' proto boards - enables you to confirm that problem persists across multiple boards as opposed to a bad connection on a single/first board. Recheck continuity between every JTAG pin - especially @ the micro pins themselves. I only have got one board. I think I do not understand your text exactly... I checked the configuration twice - but I can't find an error... A College of mine said, that I should change the settings to ''device 0'' oder ''device 1'' at the options in IAR Kickstart. What does he mean? Device 0 oder 1 should be written on the JTAG - but where I can find the options? Thanks and sorry for my bad english ;)

[jj]

Posted on May 17, 2011 at 13:10

Thank you - find your English fine.

As you can successfully ''JTAG connect'' to a development board I suspect that you have either a broken, or crossed connection between your ''custom'' board's JTAG connector and the JTAG pins @ the STM32. Be sure to probe between the JTAG connector (which accepts the JLINK) and the physical pins of the STM32. Also - I believe that it's safest for you to add pull-up resistors on each of the JTAG pins. (Suggest that you copy what has been done on your working development board - as regards JTAG pins & connections) Last - wonder if you have a Reset problem @ the STM32? Again check that your custom board agrees with your development board. If all of the above does not restore JTAG I suggest that you build a 2nd pcb - and install a fresh STM32 - to eliminate a ''single board'' issue. Good luck.

[rael]

Posted on May 17, 2011 at 13:10

If the MCU's not being held in reset, and if your JTAG wiring is correct, then perhaps you have some unconnected VSS and VDD pins on the chip ? Some people assume that there are low resistance connections internally, but you need to connect each pin externally.

[anm]

Posted on May 17, 2011 at 13:10

Quote:

On 02-05-2009 at 14:47, Anonymous wrote: Thank you - find your English fine. As you can successfully ''JTAG connect'' to a development board I suspect that you have either a broken, or crossed connection between your ''custom'' board's JTAG connector and the JTAG pins @ the STM32. Be sure to probe between the JTAG connector (which accepts the JLINK) and the physical pins of the STM32. So, I checked it again, I took a small cable to get the pins of the stm32 - so the connection should be correct. The other side, I took the connector of the JLINK (inclusive cabling).

Quote:

Also - I believe that it's safest for you to add pull-up resistors on each of the JTAG pins. (Suggest that you copy what has been done on your working development board - as regards JTAG pins & connections) I have done that already. Here is the schematics: http://testaccount.ywh-server.info/schematics.pdf

Quote:

Last - wonder if you have a Reset problem @ the STM32? Again check that your custom board agrees with your development board. I'm going to test this on monday - then I have time to check it.

Quote:

If all of the above does not restore JTAG I suggest that you build a 2nd pcb - and install a fresh STM32 - to eliminate a ''single board'' issue. I changed the STM32 already, because of an other issue. A second pcb will come in about one to two weeks - then I could test it there.

Quote:

Good luck. Thanks! And many thanks for your help - I wish you a nice sunday. EDIT: My PDF-Printer has some problems with straigt lines, so the connections to the pins are really there...[ This message was edited by: anm on 02-05-2009 16:39 ]

[anm]

Posted on May 17, 2011 at 13:10

Hi!

I will check if the controller is hold in reset - tomorrow. The VDD and VCC Pins are all externally connected... Thanks. Andi

[anm]

Posted on May 17, 2011 at 13:10

Hi!

All pins have VCC/VDD correctly, the JTAG Pins are connected correctly and the controller is not hold in reset (NRST is high). What else could that be? thanks

[jj]

Posted on May 17, 2011 at 13:10

As you have at least one, working development board suggest that you scope probe & capature the waveforms @ each STM32 JTAG pin as the JTAG connection is first established. Then repeat the procedure with your custom board. It should be that ''some'' JTAG data runs - for some period - even on your failing, custom board.

Insure that the JTAG signal transitions clearly qualify as ''hi'' and ''lo.'' Look especially at the JTAG clock signal. Also - couldn't hurt for you to ''s l o w'' your JTAG data rate - which you should be able to do w/in IAR and/or one of the Segger applications. You do have an identical HF crystal on your custom board? Can you borrow the xtal from working dev. board - just in case? My firm's not encountered this issue - across hundreds of STM32 (and brand X) similar devices. After trying this why not email Segger - describing all of the things we've suggested. Appreciate hearing ''how'' you finally correct/solve this...