look for a methold to solve the problem about GC2011A-PB

Hello,

The GC2011A function can be tested in several methods.   There are several differnent filtering modes, it is important to connect the SI to some specific time relationship of the input, and with the external output include the SO and DAV functions.  You can test the physical interfaces with the fullrate test first.  If you want to use other modes, the suggestion is to test the fullrate mode, and then select and test other modes.  

If we divide the testing into 5 sections, you can test

   a) the GC2011A register control,

   b) snapshot memory read of the counter input

   c) snapshot capture of the Filter input,

   d) snapshot capture of the filter output,

   e) and with your external circuit testing a filtered function using the snapshot capture of the input, creating a full rate filtered version in software, and comparing the software   result with the snapshot capture of the filter output.

The GC2011A has several operating modes, the easiest one to verify is a fullrate input, and fullrate output.

Section 3.1 in the datasheet lists the different options for the filter in the first column, if the filter has a real or complex input - column2, and the programming values for some of the registers in columns 4 - 9.

Since the coefficients can be stored in one of 2 banks, you need to select the bank for the coefficient memory and program the coefficients and bank selector.

a) section 3.10 lists the diagnostic test - the suggestion is to program the registers with the table 11 values, and read back the registers, and snapshot memory.  If you get the correct answer, then the register control portion works.

b) section 2.14 discusses the snapshot memory, it can be selected to 16bit 256 word mode, to collect the input or output.  You will need to write software to configure, trigger and retrieve the snapshot memory.  In the first test, configure the 2011A for full rate, non symmetric coefficients.  Use an impulse tap as the first of the 16 taps.  Enable the counter input mode.  Collect a snapshot of the output.  You should collect a ramp (counter pattern), scaling the coefficient and output gain will help to determine if this mechanism is working. 

 c) depending on the datasource for the GC2011A input - with the same scaled impulse, and full rate, input a data pattern from the signal source (could be a ramp) or inband low frequency tone.  Collect the snapshot capture of the filter input you want to make sure that all of the data bits are toggling, and that offset binary or 2s complement negative and positive values are tested.

d) once (c) is working, the snapshot capture can move to the filter output, you can externally process in matlab or C, the convolved input with a FIR filter, and a scaled output.  You can compare the captured output with the software output.

e) by having a snapshot memory collection (even if it is a temporary test function) in your digital receive circuit, you can repeat (d) and include the filter output pins.

Regards,

Radio Joe

hello,

Thank you very much for your reply. the testing methold is excellent  . according to  your  suggestion  , the  detailing  procedures   are  satifying  .

 Expecting   the farther  conmunication  with you !

                  regards,

                  kongwei

Hello,

The GC2011A has several different modes to provide filtering.  The coefficients for the filter and the selected data path signal number of bits are important to the analysis.

Please send the GC2011A mode selected, and the filtering requirements if you would like assistance.  In some cases the GC5016 can be used as an alternate part for filtering.

Regards,

Radio Joe