aboutsummaryrefslogtreecommitdiffstats
path: root/fpga/usrp3/lib/dsp/duc_chain.v
blob: 44931318c9954f0f442c771e3ee6389e5409df9c (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
//
// Copyright 2011-2013 Ettus Research LLC
// Copyright 2018 Ettus Research, a National Instruments Company
//
// SPDX-License-Identifier: LGPL-3.0-or-later
//


//! The USRP digital up-conversion chain

module duc_chain
  #(
    parameter BASE = 0,
    parameter DSPNO = 0,
    parameter WIDTH = 24,
    parameter NEW_HB_INTERP = 0,
    parameter DEVICE = "7SERIES"
  )
  (input clk, input rst, input clr,
   input set_stb, input [7:0] set_addr, input [31:0] set_data,

   // To TX frontend
   output [WIDTH-1:0] tx_fe_i,
   output [WIDTH-1:0] tx_fe_q,

   // From TX control
   input [31:0] sample,
   input run,
   output strobe,
   output [31:0] debug
   );

   genvar 	 i;


   wire [17:0] scale_factor;
   wire [31:0] phase_inc;
   reg [31:0]  phase;
   wire [7:0]  interp_rate;
   wire [3:0]  tx_femux_a, tx_femux_b;
   wire        enable_hb1, enable_hb2;
   wire        rate_change;

   setting_reg #(.my_addr(BASE+0)) sr_0
     (.clk(clk),.rst(rst),.strobe(set_stb),.addr(set_addr),
      .in(set_data),.out(phase_inc),.changed());

   setting_reg #(.my_addr(BASE+1), .width(18)) sr_1
     (.clk(clk),.rst(rst),.strobe(set_stb),.addr(set_addr),
      .in(set_data),.out(scale_factor),.changed());

   setting_reg #(.my_addr(BASE+2), .width(10)) sr_2
     (.clk(clk),.rst(rst),.strobe(set_stb),.addr(set_addr),
      .in(set_data),.out({enable_hb1, enable_hb2, interp_rate}),.changed(rate_change));

   // Strobes are all now delayed by 1 cycle for timing reasons
   wire        strobe_cic_pre, strobe_hb1_pre, strobe_hb2_pre;
   reg 	       strobe_cic = 1;
   reg 	       strobe_hb1 = 1;
   reg 	       strobe_hb2 = 1;

  assign strobe = strobe_hb1;

   cic_strober #(.WIDTH(8))
     cic_strober(.clock(clk),.reset(rst),.enable(run & ~rate_change),.rate(interp_rate),
		 .strobe_fast(1'b1),.strobe_slow(strobe_cic_pre) );
   cic_strober #(.WIDTH(2))
     hb2_strober(.clock(clk),.reset(rst),.enable(run & ~rate_change),.rate(enable_hb2 ? 2'd2 : 2'd1),
		 .strobe_fast(strobe_cic_pre),.strobe_slow(strobe_hb2_pre) );
   cic_strober #(.WIDTH(2))
     hb1_strober(.clock(clk),.reset(rst),.enable(run & ~rate_change),.rate(enable_hb1 ? 2'd2 : 2'd1),
		 .strobe_fast(strobe_hb2_pre),.strobe_slow(strobe_hb1_pre) );

   always @(posedge clk) strobe_hb1 <= strobe_hb1_pre;
   always @(posedge clk) strobe_hb2 <= strobe_hb2_pre;
   always @(posedge clk) strobe_cic <= strobe_cic_pre;

   // NCO
   always @(posedge clk)
     if(rst)
       phase <= 0;
     else if(~run)
       phase <= 0;
     else
       phase <= phase + phase_inc;

   wire        signed [17:0] da, db;
   wire        signed [35:0] prod_i, prod_q;

   wire [17:0] i_interp, q_interp;

   wire [17:0] hb1_i, hb1_q, hb2_i, hb2_q;

   wire [7:0]  cpo = enable_hb2 ? ({interp_rate,1'b0}) : interp_rate;
   // Note that max CIC rate is 128, which would give an overflow on cpo if enable_hb2 is true,
   //   but the default case inside hb_interp handles this
   generate
      if (NEW_HB_INTERP == 1) begin: new_hb
	 // First stage of halfband interpolation filters. These run at a max CPO of 2 when CIC is bypassed and HB2 enabled.
	 hb47_int
	   #(.WIDTH(18),
	     .DEVICE(DEVICE))
	     hb1_i0
	       (
		.clk(clk),
		.rst(rst),
		.bypass(~enable_hb1),
		.stb_in(strobe_hb1),
		.data_in({sample[31:16],2'b00}),
		.output_rate(cpo),
		.stb_out(strobe_hb2),
		.data_out(hb1_i)
		);

	 hb47_int
	   #(.WIDTH(18),
	     .DEVICE(DEVICE))
	     hb1_q0
	       (
		.clk(clk),
		.rst(rst),
		.bypass(~enable_hb1),
		.stb_in(strobe_hb1),
		.data_in({sample[15:0],2'b00}),
		.output_rate(cpo),
		.stb_out(strobe_hb2),
		.data_out(hb1_q)
		);

	 // Second stage of halfband interpolation filters. These run at a max CPO of 1 when CIC is bypassed.
	 hb47_int
	   #(.WIDTH(18),
	     .DEVICE(DEVICE))
	     hb2_i0
	       (
		.clk(clk),
		.rst(rst),
		.bypass(~enable_hb2),
		.stb_in(strobe_hb2),
		.data_in(hb1_i),
		.output_rate(interp_rate),
		.stb_out(strobe_cic),
		.data_out(hb2_i)
		);

	 hb47_int
	   #(.WIDTH(18),
	     .DEVICE(DEVICE))
	     hb2_q0
	       (
		.clk(clk),
		.rst(rst),
		.bypass(~enable_hb2),
		.stb_in(strobe_hb2),
		.data_in(hb1_q),
		.output_rate(interp_rate),
		.stb_out(strobe_cic),
		.data_out(hb2_q)
		);

      end else begin: old_hb

	 hb_interp #(.IWIDTH(18),.OWIDTH(18),.ACCWIDTH(WIDTH)) hb_interp_i
	   (.clk(clk),.rst(rst),.bypass(~enable_hb1),.cpo(cpo),.stb_in(strobe_hb1),.data_in({sample[31:16], 2'b0}),.stb_out(strobe_hb2),.data_out(hb1_i));
	 hb_interp #(.IWIDTH(18),.OWIDTH(18),.ACCWIDTH(WIDTH)) hb_interp_q
	   (.clk(clk),.rst(rst),.bypass(~enable_hb1),.cpo(cpo),.stb_in(strobe_hb1),.data_in({sample[15:0], 2'b0}),.stb_out(strobe_hb2),.data_out(hb1_q));

	 small_hb_int #(.WIDTH(18)) small_hb_interp_i
	   (.clk(clk),.rst(rst),.bypass(~enable_hb2),.stb_in(strobe_hb2),.data_in(hb1_i),
	    .output_rate(interp_rate),.stb_out(strobe_cic),.data_out(hb2_i));
	 small_hb_int #(.WIDTH(18)) small_hb_interp_q
	   (.clk(clk),.rst(rst),.bypass(~enable_hb2),.stb_in(strobe_hb2),.data_in(hb1_q),
	    .output_rate(interp_rate),.stb_out(strobe_cic),.data_out(hb2_q));

      end // block: old_hb
   endgenerate

   cic_interp  #(.bw(18),.N(4),.log2_of_max_rate(7))
     cic_interp_i(.clock(clk),.reset(rst),.enable(run & ~rate_change),.rate(interp_rate),
		  .strobe_in(strobe_cic),.strobe_out(1'd1),
		  .signal_in(hb2_i),.signal_out(i_interp));

   cic_interp  #(.bw(18),.N(4),.log2_of_max_rate(7))
     cic_interp_q(.clock(clk),.reset(rst),.enable(run & ~rate_change),.rate(interp_rate),
		  .strobe_in(strobe_cic),.strobe_out(1'd1),
		  .signal_in(hb2_q),.signal_out(q_interp));

   localparam  cwidth = WIDTH;  // was 18
   localparam  zwidth = 24;  // was 16

   wire [cwidth-1:0] da_c, db_c;
   //
   // Note. No head room has been added to the CORDIC to accomodate gain in excess of the input signals dynamic range.
   // The CORDIC has algorithmic gain of 1.647, implementation gain of 0.5 and potential gain associated with rotation of 1.414.
   // Thus the CORDIC will overflow when rotating and an input CW with (clipped) effective amplitude of 1.22 is applied.
   //
   cordic_z24 #(.bitwidth(cwidth))
     cordic(.clock(clk), .reset(rst), .enable(run),
	    .xi({i_interp,{(cwidth-18){1'b0}}}),.yi({q_interp,{(cwidth-18){1'b0}}}),
	    .zi(phase[31:32-zwidth]),
	    .xo(da_c),.yo(db_c),.zo() );

   MULT_MACRO #(.DEVICE(DEVICE),  // Target Device: "VIRTEX5", "VIRTEX6", "SPARTAN6","7SERIES"
		.LATENCY(1),         // Desired clock cycle latency, 0-4
		.WIDTH_A(18),        // Multiplier A-input bus width, 1-25
		.WIDTH_B(18))        // Multiplier B-input bus width, 1-18
   mult_i (.P(prod_i),             // Multiplier output bus, width determined by WIDTH_P parameter
	   .A(da_c[cwidth-1:cwidth-18]),// Multiplier input A bus, width determined by WIDTH_A parameter
	   .B(scale_factor),       // Multiplier input B bus, width determined by WIDTH_B parameter
	   .CE(1'b1),              // 1-bit active high input clock enable
	   .CLK(clk),              // 1-bit positive edge clock input
	   .RST(rst));             // 1-bit input active high reset

   MULT_MACRO #(.DEVICE(DEVICE),  // Target Device: "VIRTEX5", "VIRTEX6", "SPARTAN6","7SERIES"
		.LATENCY(1),         // Desired clock cycle latency, 0-4
		.WIDTH_A(18),        // Multiplier A-input bus width, 1-25
		.WIDTH_B(18))        // Multiplier B-input bus width, 1-18
   mult_q (.P(prod_q),             // Multiplier output bus, width determined by WIDTH_P parameter
	   .A(db_c[cwidth-1:cwidth-18]),// Multiplier input A bus, width determined by WIDTH_A parameter
	   .B(scale_factor),       // Multiplier input B bus, width determined by WIDTH_B parameter
	   .CE(1'b1),              // 1-bit active high input clock enable
	   .CLK(clk),              // 1-bit positive edge clock input
	   .RST(rst));             // 1-bit input active high reset


   wire [32:0] 	     i_clip, q_clip;

   // Cordic rotation coupled with a saturated input signal can cause overflow
   // so we clip here rather than allow a wrap.
   clip_reg #(.bits_in(36), .bits_out(33), .STROBED(1)) clip_i
     (.clk(clk), .in(prod_i[35:0]), .strobe_in(1'b1), .out(i_clip), .strobe_out());
   clip_reg #(.bits_in(36), .bits_out(33), .STROBED(1)) clip_q
     (.clk(clk), .in(prod_q[35:0]), .strobe_in(1'b1), .out(q_clip), .strobe_out());

   assign tx_fe_i = i_clip[32:33-WIDTH];
   assign tx_fe_q = q_clip[32:33-WIDTH];


   //
   // Debug
   //
   assign 	     debug = {strobe_cic, strobe_hb1, strobe_hb2,run};

endmodule // duc_chain