F28335 Held in Reset?

I met the same problem that CPU held on reset status.

 i designed a PCB with f28035. The CPU is TMX320F28035PN got from TI's sample, the power and  crystal circuit are typical usage. a 20MHz crystal attached to the on-chip oscillator circuit. Internal VREG on(/VREGENZ pin is tied low). /XRS connects to power 3.3V with 10k Resistance. When power on, checking VDD is 1.9V. VDDIO is 3.3V. It can't be connected to CCS4.0 through XDS100 Emulator. Checking /XRS pin by oscilloscope, a low pulse appears continualy. The interval between two pulses is about 13ms, pluse width is 50us.

From datasheet i know that /XRS outputs reset pulse maybe:

1. The NMI-Watchdog engages in case of a clock failure and can generate a device reset

2. When the on-chip VREG is used, its output is monitored by the POR/BOR circuit, which will reset the device should the core voltage (VDD) go out of range.

Is there any other reasons to drive /XRS reset ?  How to get the reason of /XRS reset? Is this CPU damaged?

TO Greg Stephenson: I sugest you checking the /XRS pin. Maybe it's in the same  status

Is there anyone meet this situation?

thanks

xuhong

I've got the same problem.

My F28334 runs with a 33.333MHz crystal (internal with 100MHz) and it worked fine.
Suddenly the controller didn't start at all.
The reset pin xrs pulses with 4ms intervall and 16us pulse wide to low.

I don't have any emulator access to the controller.

Is the controller damaged ?

thanks
Mike

Hi Greg,

I had the same problem ( http://e2e.ti.com/support/microcontrollers/tms320c2000_32-bit_real-time_mcus/f/171/p/21578/83389.aspx?PageIndex=2 ) with my ezdsp28335 dsp and i was unable to find some answer to this, maybe ( i´m guessing ) is a chip problem. I tried to reset the emulator, re-installing the CCS 3.3.

Maybe you have more chance to get an answer.

Regards

Gastón

To those with the same problem,

I wound up replacing the chip, I believe something fried. After experimenting I believe the cause of my problem was the ADC unit generating excess heat. If no pre-scaler was applied to the ADC, the chip grew quite hot within 30-40 seconds (this is the replacement), however by slowing the ADC clock down to 25 MHz I was able to run the chip successfully for a few minutes at a time. Replacing the Kit base with the same control card did not change the outcome. Even kept cool the ADC was far less reliable than a $3 AVR ATMEGA48 I was using to check my setup. I haven't heard back from TI either in response to the emails I sent or on the forums in near on 2 months now, so unless you represent significant purchase potential, don't expect support.

Hi everybody

I changed the boot selection from FLASH to SCI-A Boot.
The periodic reset pulse vanished and I hoped to get access to the controller with the emulator. It seemed to get normal again.
But now, the emulator couldn't erase the FLASH. "Failed to unlock" with FFFF password. I didin't  use a password.

I think the FLASH memory (and the password section) is completely disordered for unknown reasons.

A damaged chip or a unknown overwritten password is the same at the end. => Replace

Regards
Mike

Hi again

The same problem again.
I replaced the "damaged" F28334. It worked so far. I made a few programings with the flash programmer and the USB2000 emulator. 
But suddenly the controller starts to reset with a 4ms intervall and 16us pulse wide.
It began WITHOUT a new flash programing !
I changed the boot selection to have emulator access. After that the flash programmer has come with a  

"Flash Programmer: Error erasing flash memory. Device is locked or not connected"

Now, I think there isn't a overwritten password section. I didn't program the flash. The fault apeared suddenly after a few good restarts !

Does anybody have an idea ?

Regards
Mike

MIke,

Check the year of my post, since then I don't have any answer.  What  I´ve done is replace the USB cable to a shielded one, it worked for a few minutes, maybe that will give you enough time to make some changes.

Regards

Gastón

Hi Gaston

Thank you for your advise. It didn't help.

I tried with the depletion recovery button in the on-chip-flash window. It didn't help either.
I also tried the unlock button in the on-chip-flash window. It didn't work. CCS crashed !

I use CCS 4.2.0.10017 with the newest updates and a BH USB2000 emulator.

Regards
Mike

Hi Mike,

I guess it would be glad if someone from Texas can explain about this issue that is a common thing ( check the forum ) in this kind of board ( using the tms320f28335 processor ), buying an external emulator ( u$s 500 more or less ) it won´t assure you if the board or the processor is working. As you can see I´ve been trying to find an straight answer to this issue without hope.  Right now i have a nice paperweight in my home. I bought an F28027 to work because is less expensive  and if you have some problems ( compare to an u$s 500 board - ezdsp28335 -  like I have ) you lose almost nothing.

Regards

Gastón

Mike said:

I replaced the "damaged" F28334. It worked so far. I made a few programings with the flash programmer and the USB2000 emulator. 
But suddenly the controller starts to reset with a 4ms intervall and 16us pulse wide.
It began WITHOUT a new flash programing !
I changed the boot selection to have emulator access. After that the flash programmer has come with a  

"Flash Programmer: Error erasing flash memory. Device is locked or not connected"

Now, I think there isn't a overwritten password section. I didn't program the flash. The fault apeared suddenly after a few good restarts !

Mike,

I would suggest checking all of the hardware connections.  Is that a socketed eZdsp?  I've seen issues where the device was not seated well in the socket or became loose and wasn't making a good connection.   If, in particular, the flash power is not good the flash will read back 0's which results in a locked condition.

-Lori

Lori,

Lori said:

I would suggest checking all of the hardware connections.  Is that a socketed eZdsp?  I've seen issues where the device was not seated well in the socket or became loose and wasn't making a good connection.   If, in particular, the flash power is not good the flash will read back 0's which results in a locked condition.

I checked all the hardware connection too, I've replaced the USB cable ( by a shielded one ), I´ve tried to boot from FLASH,  I've tried to make a software reset  and a hardware reset. I contact to tecnical support and they recommend to erase all the register of the machine before re-install code composer. 

Maybe ( i'm guessing ) there is something wrong with the chip, because if you check the forum you´ll see post like this ( with issues like this ). My problem (like this one) was last year and I just got one reply from just one guy , since then, I see post like this increaseing. Why??

Thanks Lori.

Gastón

Gaston said:

I checked all the hardware connection too, I've replaced the USB cable ( by a shielded one ), I´ve tried to boot from FLASH,  I've tried to make a software reset  and a hardware reset. I contact to tecnical support and they recommend to erase all the register of the machine before re-install code composer. 

Thank you for the feedback Gaston.

I think there are a couple of issues that I am seeing.  First being able to connect to the device at all and second, once connected an issue with the flash appearing to be locked.

You mentioned that you tried boot to flash - can you also try "boot to check mode".  This will make sure the boot code stays within the ROM and doesn't trip any of the security mechanisms while CCS tries to connect.  It will also keep the resets from happening that may also be locking out the device.

Unfortunately the only way to try to debug this is to work through the individual potential causes.  There may indeed be damage to the particular device at this point. 

We have a couple of wiki pages that talk about debugging JTAG issues - it sounds like you may have already gone through these but if not they may help:

http://processors.wiki.ti.com/index.php/Troubleshooting_C2000_Debugger_Connection_Issues

http://processors.wiki.ti.com/index.php/Debugging_JTAG_Connectivity_Problems

http://processors.wiki.ti.com/index.php/Code_Security_Module_FAQ_for_C2000#Reasons_CSM_Appears_Secure

Regards,

Lori

-Lori

Any luck with this, Gaston and Mike?

-Lori

Lori,

Thanks for your interest, but in my case is not working at all from last year ( not only tried to boot from flash, also from check mode and any other possibilty that i could have). I re-check the hardware. Sad because I was doing my thesis with this hardware and it forces me to change not only the processor. The department of my faculty was interested in buy this hardware but this kind of problem makes you re-think the inversion.

I was thinking in buy the external emulator to work with my board, but I guess the inversion ( $ 500 ) it won't assure if the board is okay. Check the link i post in this tread and you'll see the year of my problem.

Thanks anyway Lori.

Gastón

Hi!

Is this about the same problem I have?

Failed Software Reset:
Error 0x00000024/-1137
Error during: Register, Execution,
It appears that the target is being held in reset.  This may be
due to Wait-In-Reset (WIR) configuration set by the EMU 0/1 pin
settings.  If this is the case, press Cancel and disconnect all
connected devices to attempt to clear the WIR configuration.

This is the message I get when I try to load the program. Everything worked fine so far. I don't know what to do with this, I can't program the board anymore...

The LED 2 stays on. I'm F28069.

Does anybody know something about this? I mean, am I wasting my time here, better buy a new one? And if yes, how often should I expect this?

Thanks!

Monica

I have the exact same error message, using an F28334 with my own board. CCS gives me an option to do a "Rude Attempt" to force connect, and that works and allows me to access the chip again. I hope this is not a chip hardware fault. TI, any feedback?

Servus Cristian!

CCS gives me just the options Cancel and Disconnect. None of these work... No answer from TI so far... I'm still waiting today, then I'm thinking to start making a ticket with this issue every 30min, maybe they will answer...

Thanks anyway!

Monica

spor la treaba ;)

Servus si tie :)

The first dialog that appears only has Cancel and Disconnect . If you click Cancel, you will get a second dialog that has Cancel and Rude Attempt.

Cristian

Servus!

I've tried that, I don't get the second dialog. I've installed some updates and the board is connecting now, but is still held in reset. I don't get the dialog anymore, just the errors in the console... Big change what can I say, much better now...and still no answer from TI (the answer saying they are busy this week doesn't count).

Monica

Servus,

Maybe not all emulators have the brute force option. I use an XDS510 USB...

Cristian

I have the F28335 experimenters kit and I was using CCSv3.3 when I ran into the error in "Failed Software Reset". After reading the various related posts in this thread and others, I thought I would briefly explain my experience as it has not been described and may assist others.

First off, I have seen people making statements that insinuate the error is specifically caused by software/driver problems, the emulator, or the control card. I want to clarrify that my research and experience indicates that each case could cause the error, and so assuming the cause is specific to just one of those cases is misinformation. 

My problem was obviously hardware related. I was using the experimenter kit to control a L293 based coil driving circuit and I was using the experimenter board's 3.3v headers for enable/logic signals and multiple PWM's to control drive the outputs. I took all the necessary circuit precautions to protect the DSP board from the L293 circuit, but I was working late one night, and I accidentally dropped a tool onto the L293 circuit that caused a brief short. At the time I was using VisSim to interact with the control card, and the session was ended as the experimenter board disconnected from the computer. I noticed that the L293 circuits logic LED was dimmer then usual and became concerned. I suspected that was it for my dsp board and/or the experimenter board, but then the green LEDs on both the control card and experimenter board lit back up again, and the card successfully reconnected to the PC's USB hub.

To my surprise, I noticed the DSP did not feel abnormally hot or warm and the experimenter board did not seem to have any thermal abnormalities either. Despite the DSP reconnecting, I restarted the PC just in case the hub shorted and Windows was confused of its true state, and I power cycled the experimenter board. Everything seemed fine in the device manager. Upon trying to use the experimenter board with VisSim, the standard hardware driver error that VisSim reports popped up. I opened up CCSv3.3, and attempted to connect to the board, and that's when I got the reset error. Oddly enough, it seemed as though the forced connection method worked for some of the debugging features, but a software reset would not work. It was time to bring out the multimeter.

I checked the 3.3v experimenter board header and found they measured just under 2.4v, and so I knew their was a problem. Alternatively, the control card's 3.3v test points indicated the correct 3.3v logic levels. I was relived, for this indicated that the control card, the expensive component, was not likely the cause of the error. That meant the experimenter board had a problem, but since it connected fine over USB, I was a bit surprised. Turns out, the 5v to 3.3v regulator, tps79533, was the source of the 2.4v, and despite the fact that it didn't have any visible abnormalities, it was definitely busted. From a brief datasheet glance, the  tps79533 seemed to me like a standard 3.3v regulator with additional enable and noise reduction features. I went into my scrap bin and salvaged an lm1117-3.3 surface mount regulator. The geometry, max current, and other performance features were a bit different than what the TPS had, but I figured it would suffice for a brief test of the fix and status of the control card.

Everything worked out great. The regulator was the issue, the error is gone, and the Control Card is seemingly unharmed by the incident. I intend to order a new tps79533 regulator (or get a sample if available in the appropriate package). I hope this helps anyone with a related issue. As well, I hope this encourages people to do a proper electrical check when debugging.

hey lori

i have the same problem i use ccs 5.1 to program my f28335 for some days all of the sudden the ccs gives

C28xx: Flash Programmer: Warning: The configured device (TMS320F28335), does not match the detected device (). Flash Programming operations could be affected. Please consider modifying your target configuration file.
C28xx: Failed Software Reset: (Error -1137 @ 0x0) Device is held in reset. Take the device out of reset, and retry the operation. (Emulation package 5.0.520.0)
C28xx: Trouble Reading PC Register: (Error -1137 @ 0x0) Device is held in reset. Take the device out of reset, and retry the operation. (Emulation package 5.0.520.0)
C28xx: Trouble Reading Register ST1: (Error -1137 @ 0x6) Device is held in reset. Take the device out of reset, and retry the operation. (Emulation package 5.0.520.0)
C28xx: GEL: Error while executing OnReset(1): target access failed.
C28xx: Failed Software Reset: (Error -1137 @ 0x0) Device is held in reset. Take the device out of reset, and retry the operation. (Emulation package 5.0.520.0)
C28xx: Trouble Reading PC Register: (Error -1137 @ 0x0) Device is held in reset. Take the device out of reset, and retry the operation. (Emulation package 5.0.520.0)
C28xx: Trouble Reading Register ST1: (Error -1137 @ 0x6) Device is held in reset. Take the device out of reset, and retry the operation. (Emulation package 5.0.520.0)
C28xx: GEL: Error while executing OnReset(1): target access failed.
C28xx: Trouble Writing Memory Block at 0x0 on Page 0 of Length 0x2: (Error -1137 @ 0x6) Device is held in reset. Take the device out of reset, and retry the operation. (Emulation package 5.0.520.0)
C28xx: GEL: File: C:\ti\controlSUITE\development_kits\TemplateProjects\FlashingLeds-F28335_v1.0\F2833x_RAM\FlashingLeds.out: Load failed.

i'm afraid i damage the board, if not how to take the device out of reset?

thank you

I have had the same problem (identified as wrong device when connecting) with the 28335 r2.2 control card. I noticed this problem only occurred when the device had been on for an extended period and that the micro was quite warm. I started using a fan blowing on the chip and those problems seemed to go away.

However, after leaving the system on for several days the fan was moved (blame a cluttered desk) and now the device won't even power up. I know have three control cards that have all died. I have extra experimenter kits and a 28035 control card that I have swapped so I know it's the 28335 control cards and not the emulator or software. I am using an old 5V linear supply that came with a TI DSK and my experimenters kit uses a bunch of pins but they are all ouputs. After having three boards fail, I'm about ready to try something else.

Each of the experimenters board has good 5V an 3.3V rails.

A check of the three 28335 control cards reveals:

#  D1                C28  C29

1  Dim Green  2.3V  0.3V

2   Off                 0.0V  0.0V

3   Green          2.8V   1.88V

So I think board #2 has a bad U2. Not sure what's wrong with the other two boards? Maybe they are damaged and pulling too much current.

Any advice would be appreciated. I am using a 28335 to generate inputs to test a larger embedded system. When it works it works great, but I can't afford the delays when this board fails.

It's been a long time, but if I recall correctly, in my case I blew out the 3.3v regulator on the experimenter board. I ordered a free sample from TI of the same part, replaced it, and everything returned to normal. If I were you, I would start by verifying the state of all regulators. Best of luck, and please post results for others.