DS125BR401: How to load data from external EEPROM in Master Mode?

dear Lucas,

there is only 1 DS125BR401 and the customer already pulled READ_EN  to GND. it is strange that the external EEPROM data not to be loaded at power up, but could be loaded after  reset the registers by setting the bit 6 (Reset Registers) of 0X07 register (Digital Reset and Control). would you please help to check if any action of RESET may enable EEPROM to be loaded?

I am also have have question about the function of READ_EN,  it is said " When using an External EEPROM, a transition of READ_EN pin from high to low starts LVCMOS the load from the external EEPROM", but as READ_EN is pulled to GND, will the redriver pull it high during power-up? when it pull READ_EN  to High? The customer did not notice HIGH level of READ_EN, so how could there be transition from HIGH to low?

thanks a lot!

dear Lucas,

in addition, what is the work flow of DS125BR401 to load the EEPROM data? will it keep loading the data after  READ_EN pin being pulled LOW until success or will stop loading after a certain time?

thanks! 

Hi Andrew,

Thank you for your clarifications.

Can you clarify, does EEPROM load successfully after writing 0x07[6]=1 to reset register values? Or are you writing 0x07[5]=1 to reset the SMBus master state machine?

The pin description for READ_EN is a bit misleading. EEPROM load will begin when READ_EN is pulled low, so tying READ_EN to GND will initiate EEPROM load on device power up.

DS125BR401 will attempt to load data from the EEPROM several times. After a fixed number of tries, I believe the redriver will stop attempting to load data. On a similar device, EEPROM load will be declared a failure after 64 attempts. I'm unclear if the number of attempts for DS125BR401 is different.

I have a few a debug questions.

1. Can you probe the ALL_DONE pin and report the voltage level both at power up and after register reset is performed?
2. Can you share the schematic for me to review?

Best,

Lucas

dear Lucas,

the schematics is as below:

as double checked with the customer, they reset the SMBus master by sending 00100001(0x21)  to 0X07, not reset the whole registers.

all_done pin voltage is not easy to check as their system is  immerse in water. but the EEPROM Read done bit is 0 when failed to load and turn to 1 when succeed to load.

 compared to their old project, where DS125BR401 may load EEPROM data without resetting SMBus master, there are 2 difference in this issued project:

1. the number of DS125BR401 changed to 1 from 2 (the schematics shows the connection of ENMSB & read_EN is correct)

2. the BMC device is not the same (there is also BMC device connected on the same SMBus of DS125BR401), is it the reason that cause the SMBus error when loading EEPROM upon power up? is there any difference of the Behavior of the SMBus state machine when RESET compared to power-up? 

DS125BR401 use SMBus rather than I2C Bus, I notice that customer might use it as I2C, so might it cause any confliction of comunication in some case?

Please help to share your valuable advice. the customer is in urgency now.

Thanks a lot! 

Hi Andrew,

Thank you for sharing the schematic, I reviewed device U1. Did you try to share a second image? If you did, I am unable to see it. Can you share a schematic for the second redriver?

I have a few questions based on the schematic.

• Can you try depopulating C617? I'm wondering if this capacitor is preventing the redriver from properly recognizing low input on the READ_EN pin.
• I see ALL_DONE is connected to the READ_EN pin of a second redriver. Can you confirm, do 2 redrivers share the same I2C bus? Will they both load their configuration from the same EEPROM?
• Is there any pull-down resistor on the READ_EN pin of the second redriver?
• Can you confirm SDA and SCL lines include pull-up resistors?

Thank you for the confirmation on which register bit is used to reset the SMBus master. This matches my understanding of how to restart EEPROM load.

You mention that the board is immersed in water. I have a few questions about this.

• Are components directly in contact with water? Water is a conductor and can damage components that are not specifically rated for underwater usage, including DS125BR401.
• Can the customer test with a system that has not been immersed in water before? I'd like to see if the issue still occurs without water immersion.

Regarding the BMC device: Is this device maintaining master control of the I2C bus? DS125BR401 needs to assume master control when it loads its configuration from the EEPROM. It will revert to an I2C slave after the load is complete.

SMBus is built on I2C and is therefore generally compatible with I2C devices. It is common that our customers connect the SMBus interface of our redrivers and retimers to other I2C devices. I think it is very unlikely there is some compatibility issue between I2C and SMBus.

If you're interested, you can review the following app note for more info about I2C/SMBus compatibility: sloa132_smbus_compatibility_with_i2c.pdf

Finally, can you share the name of the customer for my information?

Best,

Lucas