Marko- > We are booting (note, this is not the same as executing) the device from > EMIF. We have an 16bit wide flash device attached to it. > > Regarding the CLKOUT issue see advisory DL_13: > > "...SYSR does not control CLKOUT divisor when a JTAG connection to the > debugger is not present..." Ok I see it now. They spelled it CLOCKOUT so I missed it before. The System Register (SYSR) cannot be read or written without JTAG clock. SYSR does not control CLOCKOUT divisor when a JTAG connection to the debugger is not present. Writes to this register without JTAG fail to divide down clock out. Wow that's really surprising -- what did TI do, build the C55xx series with entry-level graduates? You would think that after so many successful C54xx devices that CLKOUT logic/circuitry would be a slam dunk. That's the type of error that results from lack of basic testing. Have you tried using a "helpful" JTAG clock? Just give TCK some signal, even with /TRST is held low? That might work. -Jeff > Jeff Brower wrote: > > > Marko- > > > > > >>Wa are using the F revision. All of the versions are buggy, but they > >>fixed some things in the latest versions (F). See the sillicon errata. > >> > >>We are still digging into a problem we discovered with USB. Randomly we > >>are missing transactions. We still don't know if this is a HW or SW bug. > >>The behaviour is that from a certain point we don't get any more USB > >>interrupts, until we unplug and then re-plug the device. > >> > >>In think the major bug you should be aware is that CLKOUT is out of > >>control when booting from FLASH (not via CCS and JTAG). Also there are > >>several bugs with sleep modes. > > > > > > Thanks, that is extremely valuable information. Yeah I've noticed the Rev D errata > > has 109 CPU advisories alone, not counting EMIF, DMA, etc. > > > > I will look into the CLKOUT issue -- I did not see it in the errata. Are you using > > HPI boot from Flash, or other peripheral? > > > > > > > >>marko > >> > >>Jeff Brower wrote: > >> > >>>Marko- > >>> > >>> > >>> > >>>>this may seems like a joke but it isn't. I suggest you to first read the > >>>>silicon errata of the device you will use. We have to change our design > >>>>several times due to the hardware silicon bugs in the 5509. > >>> > >>> > >>>Which version of silicon is buggy? TI shows both Rev D and F. > >>> > >>>Did you have problems with USB? > >>> > >>>-Jeff > >>> > >>> > >>> > >>>>jeanpierrepoulin wrote: > >>>> > >>>> > >>>>>Hi, > >>>>> > >>>>> > >>>>>I'm starting out in the brave new world of TI DSPs (from the simpler > >>>>>days of PIC designs), and am facing a very steep learning curve in > >>>>>front of me as I contemplate the design of my next DSP circuit. > >>>>> > >>>>>I can handle the software part because of my background, but the > >>>>>hardware appears daunting! > >>>>> > >>>>>My next circuit needs to: > >>>>>- Support SDRAM & Flash > >>>>>- Support JTAG interface > >>>>>- Read image frames from a CMOS camera > >>>>>- Read/Write to a CompactFlash card / Hard Drive > >>>>>- Communicate with an Ethernet MAC/PHY chips such as the Crystal > >>>>>C8900A. (IP Stack implemented on DSP) > >>>>> > >>>>>The TI / Spectrum Digital DSK/EVMs reference designs are extremely > >>>>>useful, but they unfortunately use CPLDs/FPGAs to provide the 'glue > >>>>>logic', chips I don't intend to use to keep cost-per-circuit low and > >>>>>keep programming/manufacturing simple. > >>>>> > >>>>>What I need is a working design that uses EMIF for memory and > >>>>>teaches me to use other CE spaces to access other peripherals such > >>>>>as the CompactFlash, MAC and camera... > >>>>> > >>>>>Any hints / suggestions / links you could offer to get me started? > >>>>> > >>>>>Many thanks!! > >>>>> > >>>>> Jean-Pierre |
|
Re: [Fwd: Re: Sources for good beginner TI DSPs hardware designs?]
Started by ●March 8, 2004
Reply by ●March 8, 20042004-03-08
We have ordered 20.000 pcs of JTAG debuggers from TI. We will integrate them in our boardso we will be able to control the CLKOUT pin :))) joking of course. No, we didn't test anything further, we simply externaly diveded the clock (D flip flop). marko Jeff Brower wrote: > Marko- >>We are booting (note, this is not the same as executing) the device from >>EMIF. We have an 16bit wide flash device attached to it. >> >>Regarding the CLKOUT issue see advisory DL_13: >> >>"...SYSR does not control CLKOUT divisor when a JTAG connection to the >>debugger is not present..." > Ok I see it now. They spelled it CLOCKOUT so I missed it before. > > The System Register (SYSR) cannot be read or written > without JTAG clock. SYSR does not control CLOCKOUT > divisor when a JTAG connection to the debugger is not > present. Writes to this register without JTAG fail to > divide down clock out. > > Wow that's really surprising -- what did TI do, build the C55xx series with > entry-level graduates? You would think that after so many successful C54xx devices > that CLKOUT logic/circuitry would be a slam dunk. That's the type of error that > results from lack of basic testing. > > Have you tried using a "helpful" JTAG clock? Just give TCK some signal, even with > /TRST is held low? That might work. > > -Jeff > >>Jeff Brower wrote: >> >> >>>Marko- >>> >>> >>> >>>>Wa are using the F revision. All of the versions are buggy, but they >>>>fixed some things in the latest versions (F). See the sillicon errata. >>>> >>>>We are still digging into a problem we discovered with USB. Randomly we >>>>are missing transactions. We still don't know if this is a HW or SW bug. >>>>The behaviour is that from a certain point we don't get any more USB >>>>interrupts, until we unplug and then re-plug the device. >>>> >>>>In think the major bug you should be aware is that CLKOUT is out of >>>>control when booting from FLASH (not via CCS and JTAG). Also there are >>>>several bugs with sleep modes. >>> >>> >>>Thanks, that is extremely valuable information. Yeah I've noticed the Rev D errata >>>has 109 CPU advisories alone, not counting EMIF, DMA, etc. >>> >>>I will look into the CLKOUT issue -- I did not see it in the errata. Are you using >>>HPI boot from Flash, or other peripheral? >>> >>> >>> >>> >>>>marko >>>> >>>>Jeff Brower wrote: >>>> >>>> >>>>>Marko- >>>>> >>>>> >>>>> >>>>> >>>>>>this may seems like a joke but it isn't. I suggest you to first read the >>>>>>silicon errata of the device you will use. We have to change our design >>>>>>several times due to the hardware silicon bugs in the 5509. >>>>> >>>>> >>>>>Which version of silicon is buggy? TI shows both Rev D and F. >>>>> >>>>>Did you have problems with USB? >>>>> >>>>>-Jeff >>>>> >>>>> >>>>> >>>>> >>>>>>jeanpierrepoulin wrote: >>>>>> >>>>>> >>>>>> >>>>>>>Hi, >>>>>>> >>>>>>> >>>>>>>I'm starting out in the brave new world of TI DSPs (from the simpler >>>>>>>days of PIC designs), and am facing a very steep learning curve in >>>>>>>front of me as I contemplate the design of my next DSP circuit. >>>>>>> >>>>>>>I can handle the software part because of my background, but the >>>>>>>hardware appears daunting! >>>>>>> >>>>>>>My next circuit needs to: >>>>>>>- Support SDRAM & Flash >>>>>>>- Support JTAG interface >>>>>>>- Read image frames from a CMOS camera >>>>>>>- Read/Write to a CompactFlash card / Hard Drive >>>>>>>- Communicate with an Ethernet MAC/PHY chips such as the Crystal >>>>>>>C8900A. (IP Stack implemented on DSP) >>>>>>> >>>>>>>The TI / Spectrum Digital DSK/EVMs reference designs are extremely >>>>>>>useful, but they unfortunately use CPLDs/FPGAs to provide the 'glue >>>>>>>logic', chips I don't intend to use to keep cost-per-circuit low and >>>>>>>keep programming/manufacturing simple. >>>>>>> >>>>>>>What I need is a working design that uses EMIF for memory and >>>>>>>teaches me to use other CE spaces to access other peripherals such >>>>>>>as the CompactFlash, MAC and camera... >>>>>>> >>>>>>>Any hints / suggestions / links you could offer to get me started? >>>>>>> >>>>>>>Many thanks!! >>>>>>> >>>>>>> Jean-Pierre > > _____________________________________ > Note: If you do a simple "reply" with your email client, only the author of this message will receive your answer. You need to do a "reply all" if you want your answer to be distributed to the entire group. > > _____________________________________ > About this discussion group: > > To Join: Send an email to > > To Post: Send an email to > > To Leave: Send an email to > > Archives: http://www.yahoogroups.com/group/c55x > > Other Groups: http://www.dsprelated.com > > Yahoo! Groups Links > > |