Aaron- > Thanks, I've sorted it out now Ok sounds good Aaron. Good luck with your project! -Jeff > Quoting Jeff Brower <jbrower@jbro...>: > >> Aaron- >> >> > Yes, I am using the 5416. I'm afraid I must not understand how I've >> setup >> > the memory... >> >> As Brian indicated, for C5416 data memory from 0x8000 to 0xffff is the >> same physical memory as prog memory from 0x18000 to 0x18fff. The next >> available prog memory is 0x28000 to 0x28fff. >> >> > In addition I've noticed when writing to another arbitary address, >> e.g. >> > 0x2000, I get the duplicate at 0x12000, 0x22000, 0x32000 ? >> >> This one I don't know about. I suggest to focus in on one problem at a >> time, be sure you fully understand the chip and memory map (i.e. read >> the >> data sheet closely), then move on to the next problem. >> >> -Jeff >> >> > Quoting Jeff Brower <jbrower@jbro...>: >> > >> >> Aaron- >> >> >> >> > I'm experiencing a problem where by writing to a data location, say >> >> > 0x8000, program address 0x18000 in extended memory has the same >> value >> >> > written to it. >> >> >> >> Which C54x device are you using? You might be seeing expected >> results, >> >> for >> >> example >> >> with C5416. >> >> >> >> -Jeff >> >> >> >> > I've tested this by writing arbitary values to data memory (e.g. >> >> > 0x8000, 0x5000) at the beginning of my main() function. As soon as >> >> > the write is complete I can see that 0x18000 and 0x15000 in my >> >> > program memory has also been overwritten - as if 2 memory locations >> >> > are being written at once. >> >> > >> >> > This causes the expected crashing in the product I'm working on. If >> I >> >> > move all program code out of 0x10000 to 0x20000 everything works >> fine >> >> > but this is a huge waste >> >> > of memory! >> >> > >> >> > I've run into the same problem on 3 different boards so it's not a >> >> > damaged dsp. >> >> > >> >> > Parameters of interest are: >> >> > MP/MC = 0 >> >> > OVLY = 1 >> >> > DROM = 1 >> >> > I've played with these and the problem remains. >> >> > >> >> > Any help would be appreciated as this problem really has be >> stumped. >> >> > >> >> > Thanks >> >> > Aaron >> >> >> > >> > >> > > >