Spin Semiconductor

Hi Frank,
Does the FV-1 I2C interface support clock stretching from the slave device? I'm loading the FV-1 from a MCU and there's a delay right after the I2C read instruction from the FV-1 and the first byte of transmitted data as a result of the MCU setting up the transfer. I use the MCU's clock stretching mode to hold-off SCL until the data is ready to transmit. The 1st byte is transmitted but I get no ACK from the FV-1 - all subsequent bytes are not acknowledged either.

Nope, it doesn't support clock stretching. Sorry, I think the only solution will be to set up the buffer for transfer then do the program change to cause the FV-1 to start the load.

Thanks Frank!

I'm re-writing the firmware to insure the data is ready. I'll let you know how it goes.

Hi,

has someone already get it successfully, to load the firmware via i2c from a MCU?

I've been trying all day with a atmega8 and hardware i2c.
Unfortunately, so far without success

Greetings from Germany / Buxtehude

Hi Dirk,

since the FV-1 does access the 24LC32's I2C only when changing programs and on startup, I found it completely "legal" to wire my AVR ATmega8 controller's I2C in parallel to the EEPROM and FV-1. I just have to interrupt my own I2C transfers for about 100 ms after bootup and for about 30 ms when someone changes the switch bits. This "someone" is my controller itself, so it never runs into bus collision.

The routing is:
Switches --> IN [µC] OUT --> FV-1 S0..S2

You have to insert a 5V to 3,3V translator (schottky diode BAT41 and BS170 MOSFET for a simple solution) if useing the ATmega with 5V.

I made up a USB or serial XMODEM-128 upload solution for the 8 external EEPROM programs which works fine.