Use the PDI REPEAT instruction in the PDI programmer code to reduce protocol overhead and greatly improve transfer throughput. Switch bit-bang USART in the AVRISP project to Timer 1, so that Timer 0 can be used for hardware timeouts while waiting for the NVM bus or controller to become ready.

1 files changed, 2 insertions, 8 deletions

diff --git a/Projects/AVRISP/Lib/NVMTarget.h b/Projects/AVRISP/Lib/NVMTarget.h
index bd9296f74..7fde84538 100644
--- a/Projects/AVRISP/Lib/NVMTarget.h
+++ b/Projects/AVRISP/Lib/NVMTarget.h
@@ -56,12 +56,7 @@
 		#endif

 

 	/* Defines: */

-		#define FLASH_BASE                     0x00800000

-		#define EPPROM_BASE                    0x008C0000

-		#define FUSE_BASE                      0x008F0020

-		#define DATAMEM_BASE                   0x01000000

-		#define PROD_SIGNATURE_BASE            0x008E0200

-		#define USER_SIGNATURE_BASE            0x008E0400

+		#define NVM_BUSY_TIMEOUT_MS            200

 		

 		#define NVM_REG_ADDR0                  0x00

 		#define NVM_REG_ADDR1                  0x01

@@ -113,8 +108,7 @@
 	/* Function Prototypes: */

 		void     NVMTarget_SendNVMRegAddress(uint8_t Register);

 		void     NVMTarget_SendAddress(uint32_t AbsoluteAddress);

-		bool     NVMTarget_WaitWhileNVMBusBusy(void);

-		void     NVMTarget_WaitWhileNVMControllerBusy(void);

+		bool     NVMTarget_WaitWhileNVMControllerBusy(void);

 		uint32_t NVMTarget_GetMemoryCRC(uint8_t MemoryCommand);

 		void     NVMTarget_ReadMemory(uint32_t ReadAddress, uint8_t* ReadBuffer, uint16_t ReadSize);

 		void     NVMTarget_EraseMemory(uint8_t EraseCommand, uint32_t Address);