Faults in EVAL AD2S1210

tpsamuel,

Can you tell me what the value in the fault register is?  This information will help me understand the potential debug steps you should take to resolve them.   Please also indicate if this is an issue you see when the input (resolver) is at a static position or if it is rotating.

tpsamuel said:

Also want to know the use of resistors R22 and R21 which are in the paths of sin and cosine signals.

R21 and R22 are used typically to assist the device in detecting a partial or complete loss of the connectivity with the sensor.  The recommended value for these resistors is 68K ohms.  

Sean

tpsamuel,

Can you tell me what the value in the fault register is?  This information will help me understand the potential debug steps you should take to resolve them.   Please also indicate if this is an issue you see when the input (resolver) is at a static position or if it is rotating.

tpsamuel said:

Also want to know the use of resistors R22 and R21 which are in the paths of sin and cosine signals.

R21 and R22 are used typically to assist the device in detecting a partial or complete loss of the connectivity with the sensor.  The recommended value for these resistors is 68K ohms.  

Sean

The value in the fault register is 0xF0.

It is always on the board even when no resolver is connected.

tpsamuel,

For now let's focus on the case where the resolver is properly connected and energized.

Can you share more details of the input connectivity and if you've modified any of the configuration memory. I find it interesting that you have the clipping, LOS and DOS faults occurring all at the same time.

Can you also confirm your timing and procedure for clearing the fault register?   Please make sure that you are first setting the FAULT address through the configuration mode before attempting the recommended clear method description.   It isn't clear but that is a prerequisite for the clear to function correctly.

Sean

We are writing control register with 0x92 with 0x7F and excitation frequency register 0x91 with 0x0C to set excitation frequency @3KHz.

We used exactly same procedure for clearing the fault register as mentioned in datasheet. Also added the step you mentioned before clearing the fault register. Still its showing all these faults.

Okay,

1) So you are using 16-bit mode....Have you modified the RES0 and RES1 pin states to match the control register?  Typically this is only needed if A1, A0 are set to either NORMAL - POSITION or NORMAL - VELOCITY modes but it's worth making the change anyway.

2) Can you please share what your resolver outputs look line (SIN/SINLO, COS/COSLO) inputs using a scope?   This will help us determine if there is anything in particular of concern.  Please set the resolver angle to a static 45 degrees so that we can compare the relative magnitudes of SIN/COS so that we can review all of the fault indicators at once.

Sean

1) Using 16 Bit resolution. Modified the RES0 and RES1 connections also.

2)Please find the image from scope below. 

tpsamuel,

These waveforms are severely distorted and depending on what exactly I'm looking at may not have sufficient amplitude to avoid tripping the LOS threshold.   Also it appears that this waveform is centered around ground which would cause the clipping fault D7 to be set as well.    

Can you share with me how you are connecting the resolver or resolver simulator to the inputs.   That may provide some insight as to what is going wrong.

Sean