AD2S1210: An error occurs when switching from normal mode to configuration mode.

LGX,

I'm not sure I understand the source of the problem you are encountering but I would like a closer look at your timing waveforms for configuration versus normal operation to understand how you've implemented each serial pattern.   Could you also please indicate what your input stimulus looks like when the encoder output is disturbed during the switching.

Sean

Sean,

Sorry for not describing it clearly.

The timing diagram of SPI is shown in the figure.

Normal mode，A0 = 0 ，A1 = 0， /CS = 0，read position data.

Cofigure mode,  A0 = 1， A1 = 1,  /CS = 0,  write 0x80 0x81 0x82 0x83 0xF0，read FAULT POS_HI,POS_LO,VEL_HI,VEL_LO.

The resolver remains stationary. From the above test results, it can be seen that the normal mode and configuration mode read the same position data, and this data is correct.

In the above cases, the state of A0A1 remains unchanged, all zeros or all ones. When I deliberately insert a mode switch before the configuration mode, as shown in the figure below. Irregular error data appears. Everything goes wrong(Encoder A B Z, SDO data).

In fact, when only the SAMPLE signal and A1 A0 signal are sent and the SPI communication is turned off, the error still exists (Encoder A B Z).

Thanks,

LGX

LGX,

Can you please send me a simplified connection diagram/schematic for your setup?   It will be easier for me to spot anything abnormal.   I will continue to study your timing diagrams to see if I see anything obvious.

Sean

Sean,

Schematic:

LGX

LGX,

First of all my apologies for the delay in my response.  Having looked over your schematic (focusing primarily on the AD2S1210) I don't see anything immediately that jumps out.  However, could you try a simple experiment for me?   What happens to the response if you remove R194 and R201?  Does the behavior change?

Also do you see any fault indication (LOS/DOT) when the error occurs at the transition?

Sean 

LGX,

One more thing.   Try deasserting SAMPLE (SAMPLE=1) before issuing your serial commands?  Does this change the outcome?

LGX,

One more thing.   Try deasserting SAMPLE (SAMPLE=1) before issuing your serial commands?  Does this change the outcome?

LGX

Also what happens if you change the last address in your pattern from 0xF0 (Serial Reset) to 0xFF (Ready Fault).

Sean