Pressing RESET (SYSRSTI#->''low'') stops HCLK-PLL

[moedeker2]

Posted on May 27, 2003 at 17:57

Pressing RESET (SYSRSTI#->''low'') stops HCLK-PLL

[moedeker2]

Posted on May 23, 2003 at 12:25

After powerup everything works fine. My RESET-generator RN5VD (RICOH) generates an reset pulse at power of 130ms. But if I press my reset switch, which generates same pulse at SYSRETI as at powerup, HCLK-PLL stops. 318MHz crystal oscillator is running. No HCLK, PCICLK, MCLK. Power consumption goes down to 40%. I suppose that reading of the straps is the problem. But why?

Any idea's?

Bernd

[moedeker2]

Posted on May 26, 2003 at 07:19

Found the problem: I connected the strap signals DACK_ENC0-2 to an XILINX-CPLD to build DACK signals. The strap pullup value of 470K is overwritten by XILINX internal BUSHOLD-Circuity even that these pins are inputs at the CPLD. After disabling this BUSHOLD-Circuity reset works fine. YIPPI

[marios]

Posted on May 26, 2003 at 23:06

What STPC are you using? With my STPC Industrial, I sometimes see all the clock signals stop, but not upon a reset. Like if the CPU was executing forbidden code that shuts down the oscillators. I use a MAX705 for reset generation; the reset signal is clean. The only way to reset the CPU is to power down the board.

[moedeker2]

Posted on May 27, 2003 at 17:57

Sorry I forgot to say: It's an STPC Elite. But I also developed 2 boards with STPC Industrial - a long time ago. I could'nt remember such a problem like yours. Tomorow I will have a look into my old records.