XDS100 problems

Hi Eric,

What device/hardware are you using with XDS100?

http://tiexpressdsp.com/index.php/XDS100#What_hardware_and_software_is_supported.3F

A custom board with a 700MHz TMS320C6435.  The configuration was worked out by a previous TI DSP FAE.  No problems (other than being slow!) on the other boards/systems.  This one PC though just refuses to let it come up.

Eric Bryan

Eric,

You should contact the person that provided this for you, because we do not support the XDS100 for this part. Just the ones listed in the link that I sent.

Very sharp guy but he is no longer with TI.  He moved overseas.

Eric Bryan said:

Very sharp guy but he is no longer with TI.  He moved overseas.

I suspect I know whom you are referring to (Neal?), and I am somewhat suprised that he would have recommended an XDS100 on a C643x device, I am suprised that the XDS100 will even let you try to connect, as there just have not been drivers written for the XDS100 to work with devices outside of the listing on the Wiki Mariana pointed out, though there hopefully will be someday. Did he give you some driver package to allow this sort of connection? If so I would be interested in seeing it myself, as XDS100 support on the C643x device line would be great to have (though in general it is so slow I would highly recommend a XDS510 class emulator for serious development on this sort of device).

Send me an email and I will reply with the instructions he gave.  Really all he did was direct us to use the C6747 driver.   He stated that the XDS100 would work with all C6000 DSPs.

As I said earlier--this is working on several other setups we have here but not one so I suspect there is some sort of interaction depending on exactly what version some component is.  I'm trying to uninstall everything TI today and reinstall it all exactly like the other systems and see if that fixes it.

Eric Bryan

For now lets assume that this was possible somehow through a hack your sharp FAE made to enable the 6435 on the XDS100 (which is not recommended, supported, etc. by TI). Assuming this is not some bug in the emulation driver, my first suspicion would be there is something broken with the PLL2 module, I would take a look at the clock signal coming into PLL2 as well as potentially any other hardware centered around the PLL2, since this error is happening on addresses related to the PLL2 controller. It may be worth taking out the GEL file temporarily and seeing what happens when you open a view memory window on the 0x01C40C00-0x01C40FFF range and try to access it as a more controlled test method.  

Eric Bryan said:

Send me an email and I will reply with the instructions he gave.  Really all he did was direct us to use the C6747 driver.   He stated that the XDS100 would work with all C6000 DSPs.

I did not see this until I had already made my last post, this will not be necessary after doing some more research on the matter. The XDS100 might work with all C6x DSPs, however since this is entirely untested it is not suggested or supported by TI (for anyone else reading this please do not bother to try this until the support is really there), this support should be added in a future release of the XDS100.

I don't think this problem is related to the DSP processor.  I "cleaned up" a number of applications/tasks on the PC.  Startup now gets thru PLL2 and into DDR2 init.  I would bet there is something timing related.  The emulator task is being swapped out and then times out.

Eric Bryan

One thing to try would be to reduce the traffic on the USB bus.  It may not be getting enough bandwidth.

John

Hi Eric,

I am trying to connect a custom board with TMS320TCI6486 (which is having a 64x+ cpu core) using a third party XDS100 emulator. This emulator works fine for F28027 kit from texas instruments and connects using the xds100 drivers for f28xx from TI.

Can you please suggest what target configuration you use for connecting your board to CCSv3.3? I have tried using c6745/7 configuration which uses tixds6400_plus.dvr as a driver. 

Please have a look at the details of my issue here:

http://e2e.ti.com/forums/p/7760/30176.aspx#30176

Your help is higly appriciated.

Thanks,

Amit

(amitkumar.solanki@einfochips.com)

Eric Bryan said:

"Trouble Writing Memory Block at 0x1c40d00 on Page 0 of Length 0x4: Error 0x00000002/-1060 Error during: Memory,  An unknown error  prevented the emulator from accessing the processor in a timely fashion. It is recommended to RESET EMULATOR.  This will disconnect each  target from the emulator.  The targets should then be power cycled or hard reset followed by an emureset and reconnect to each target.  .."

Just to add one more possible cause to this thread, I was getting this same error earlier today (though on a DM365, CCS 4.1, and BH USB560m) and it ended up tracing down to the PLL multiplier incorrectly being set too high. This actually corresponds well to the reason you were getting the error with a low supply voltage (lower voltage often means the processor has a lower tolerance for higher frequencies), but is a bit of a different take on the problem

What did you have the PLL multiplier set to?

Steve said:

What did you have the PLL multiplier set to?

I believe it was set to a PLLM of 99, and when I arbitrarily cut that down to 50 it started working properly, I never worked out what actual frequency it ended up running at or how high I could set the PLLM before it broke down.