On 1/15/10 8:15 AM, Lars Schmidbauer wrote:
I don't think it's memory corruption being caused by the SDRAM being clocked too high because even PC-100 SDRAM is fine at 100MHz. The memory test probably specifically stresses the memory only and so it would pass. What is more likely happening is that corruption is being introduced during the calculation of decompression, and that corruption is being caused by the cpu probably. Using an FPU accuracy checker may yield a test failure but none of these packages are built to use fpu. I have code for one somewhere, which is how I figured out the FPU in my one 060 was complete trash. I wonder if there is a way to stress test all normal integer functions.oh, i forgot the boot-delay option. ST-RAM and FastRAM tests are OK! The Atari-diagnostics (CTRL ALT both shift) of ct60 only tests the ST-RAM and SRAM (DSP): OK, too. And so i reduced the clock to 66 MHz...and everything works fine now! At 90 MHz the error still occure. But i really don't like this solution- can another SDRAM work better (perhaps PC-133)?? Which SDRAM (and CT60 clock) do you use? Lars
Thanks, Mark