-9A90FD.svg){kind=icon}
DDR1-SDRAM controller with a AXI4 slave port.
Replacing SDR-SDRAM with DDR-SDRAM (DDR1) in low-end FPGA designs.
|------------------------------| |-----------|
| | | |
----->| driving clock and reset | | |
| | | |
------------| | DDR-SDRAM interface |--------->| |
| AXI4 | | | |
AXI4 master |-------->| AXI4 slave | | |
| | | | |
------------| |------------------------------| |-----------|
User Logic DDR1 Controller (ddr_sdram_ctrl.v) DDR1 chip
Many low-end FPGA development boards use SDR-SDRAM as off-chip memory, but DDR-SDRAM (DDR1) is larger and less expensive than SDR-SDRAM. And like SDR-SDRAM, DDR1 can also be directly driven by common IO pins of low-end FPGAs. I write a soft core DDR1 controller with a slave AXI4 interface for user. The features of this controller are:
- Platform Independent : Written in pure Verilog and can run on various FPGAs including Altera and Xilinx.
- Compatible : Supports DDR1 of various bit widths and capacities (this has been verify by simulation on MICRON's DDR1 models).
To demonstrate the use of this controller, I provide two demo projects:
- Self-test demo : Through the DDR1 controller, write data into DDR1, then read it out, and compare whether the read data is consistent with the written data.
- UART DDR1 read/write demo : Convert UART commands into AXI4 bus actions to read and write DDR1 through the DDR1 controller.
In addition, since the interface timing of each generation of DDR-SDRAM (such as DDR3, DDR2, DDR1) is similar, this repository can also facilitate those who are familiar with Verilog to learn the DDR-SDRAM interface.
- Hardware Design Guidelines
- Hard Design Example
- DDR1 Control Module Manual
- FPGA Demo Projects
- RTL Simulation
For FPGA selection, only an FPGA with a sufficient number of common IOs can drive DDR1. The IO level standard of DDR1 is often SSTL-2, which is compatible with 2.5V LVTTL or 2.5V LVCMOS, so the power supply of the corresponding FPGA IO bank should be 2.5V, and should be configured as 2.5V LVTTL or 2.5V LVCMOS in the FPGA development software .
The following table shows the pin defination of the DDR1 chip and the points that should be paid attention to when connecting it to the FPGA.
| pin name | direction | width | introduction | level | note |
|---|---|---|---|---|---|
| ddr_ck_p | FPGA output | 1 | DDR1 clock p , ≥75MHz | 2.5V LVTTL | differential route with ddr_ck_n |
| ddr_ck_n | FPGA output | 1 | DDR1 clock n , ≥75MHz | 2.5V LVTTL | differential route with ddr_ck_p |
| ddr_cke, ddr_cs_n | FPGA output | 1 | low speed | 2.5V LVTTL | |
| ddr_ras_n, ddr_cas_n, ddr_we_n | FPGA output | 1 | sync with ddr_ck_p's falling edge | 2.5V LVTTL | same length as ddr_ck_p |
| ddr_ba | FPGA output | 2 | sync with ddr_ck_p's falling edge | 2.5V LVTTL | same length as ddr_ck_p |
| ddr_a | FPGA output | depend on DDR1 part | sync with ddr_ck_p's falling edge | 2.5V LVTTL | same length as ddr_ck_p |
| ddr_dqs | inout | depend on DDR1 part | 2.5V LVTTL | ||
| ddr_dm、ddr_dq | inout | depend on DDR1 part | sync with ddr_dqs's double edge | 2.5V LVTTL | same length as ddr_dqs |
For demonstration, I used the cheapest FPGA of Altera Cyclone IV (model: EP4CE6E22C8N) and MICRON's 64MB DDR1 (model MT46V64M8TG-6T) to design a small board. All the demo project of this repository can run on this board directly. If you want to use DDR1 in your own PCB design, just refer to this board's design.
 |
|---|
| Figure : FPGA + DDR1 demo board. |
See board schematic PCB/sch.pdf and manufacturing file PCB/gerber.zip . It is a double-layer board, and it is not necessary to pay attention to impedance matching like DDR2 and DDR3, because the operating frequency of the circuit is 75MHz on both sides, which is not particularly high. Just pay attention to keeping the distance between FPGA and DDR as close as possible, and the wiring as short as possible. For example, I put the DDR1 chip on the opposite side of the FPGA chip to keep the wiring short.
The design of this board is open in LCEDA, see oshwhub.com/wangxuan/fpga-ddr-ce-shi-ban .
See RTL/ddr_sdram_ctrl.v for the DDR1 controller code, which can automatically initialize DDR1 and refresh it regularly. The module has a AXI4 slave interface through which reads and writes to DDR1 can be accomplished. This section introduce in detail how to use this module.
Verilog parameters of this module are defined as follows:
module ddr_sdram_ctrl #(
parameter READ_BUFFER = 1,
parameter BA_BITS = 2,
parameter ROW_BITS = 13,
parameter COL_BITS = 11,
parameter DQ_LEVEL = 1,
parameter [9:0] tREFC = 10'd256,
parameter [7:0] tW2I = 8'd7,
parameter [7:0] tR2I = 8'd7
)These parameters are described in the table below:
| Parameter | Type | value range | default value | introduction |
|---|---|---|---|---|
| READ_BUFFER | 0 or 1 | 1 | If it is set to 0, there will be no read data buffer in the DDR1 controller, and the read data will not wait for the AXI4 host to accept it, that is, the rvalid signal will not wait for the rready signal, and will be directly poured out at the highest rate. At this time, AXI4 is not Complete, but can reduce the latency of readout. If it is set to 1, there will be a large enough read data buffer in the DDR1 controller, and the rvalid signal will shake hands with the rready signal to confirm that the AXI4 host is ready before reading the data. | |
| BA_BITS | Width | 1~4 | 2 | The width of DDR BANK ADDRESS (ddr_ba) is specified. Regular DDR1 BANK ADDRESS is 2bit, so this parameter is usually fixed to the default value and does not need to be changed. |
| ROW_BITS | Width | 1~15 | 13 | It specifies the width of the DDR ROW ADDRESS, and also determines the width of the address line pins (ddr_a) of the DDR1 chip. This parameter depends on the selection of the DDR1 chip. For example, MT46V64M8 has 8192 COLs per bank, considering 2^11=8192, this parameter should be 13. Similarly, for MT46V128M8, this parameter should be 14 |
| COL_BITS | Width | 1~14 | 11 | Specifies the width of the DDR COL ADDRESS. This parameter depends on the selection of the DDR1 chip. For example, MT46V64M8 has 2048 COLs per ROW. Considering 2^11=2048, this parameter should be 11. Similarly, for MT46V32M16, this parameter should be 10. |
| DQ_LEVEL | Width | 0~7 | 1 | The data bit width of the DDR1 chip is specified. For a chip with a bit width of x4 (such as MT46V64M4), this parameter should be 0; for a chip with a bit width of x8 (such as MT46V64M8), this parameter should be 1; for a chip with a bit width of x16 ( For example, MT46V64M16), this parameter should be 2; for the case of bit width x32 (such as two pieces of MT46V64M16 extended bit width), this parameter should be 3; and so on. |
| tREFC | Timing | 1~1023 | 256 | The controller will refresh DDR1 periodically, and this parameter specifies the refresh interval of DDR1. See [Timing Parameters](#Timing Parameters) for details. |
| tW2I | Timing | 1~255 | 7 | This parameter specifies the interval from the last write command of a write action to the activation command (ACT) of the next action. See [Timing Parameters](#Timing Parameters) for details. |
| tR2I | Timing | 1~255 | 7 | This parameter specifies the interval from the last read command of a read action to the activation command (ACT) of the next action. See [Timing Parameters](#Timing Parameters) for details. |
This module requires a drive clock and a drive reset, as follows:
input wire rstn_async,
input wire drv_clk,rstn_async is a low-level reset signal and should be set high during normal operation. drv_clk is the driving clock, and its frequency is 4 times the user clock.
Before the module starts working, the rstn_async signal should be set low to reset the module, and then set rstn_async high to release the reset.
In this module, the driving clock drv_clk is divided by 4 to generate the DDR1 clock (ddr_ck_p/ddr_ck_n) and the AXI4 bus user clock (clk). This section describes how to determine the frequency of the drive clock drv_clk.
First, the clock frequency is limited by the DDR1 chip. Considering that all DDR1 interface frequencies are at least 75MHz, the lower limit of drv_clk is 75*4=300MHz.
The upper limit of drv_clk also depends on the chip model of DDR1. For example, for MT46V64M8P-5B, check the chip manual, the maximum clock frequency of DDR1 with -5B suffix is 133MHz when CAS Latency (CL)=2, then the upper limit of drv_clk is 133 *4=532MHz.
Note: This controller has a fixed CAS Latency (CL) = 2.
In addition, the upper limit of the clock frequency is also limited by the speed of the FPGA. Too high a clock frequency can easily lead to timing failure. This design fully considers the timing safety design, most registers work in the clk clock domain with a lower frequency; some registers work in a clock that is twice the frequency of the clk clock, and the combinational logic of the input port is very short; Under the drv_clk of the frequency, but the input port comes directly from the register output of the upper stage (no combinational logic). Therefore, even on the EP4CE6E22C8N with a very low speed class, the correct operation of the module is guaranteed at a drive clock of 300MHz.
Following is the DDR1 interface of this module. These signals should be pinout directly from the FPGA and connected to the DDR1 chip.
output wire ddr_ck_p, ddr_ck_n,
output wire ddr_cke,
output reg ddr_cs_n,
output reg ddr_ras_n,
output reg ddr_cas_n,
output reg ddr_we_n,
output reg [ BA_BITS-1:0] ddr_ba,
output reg [ ROW_BITS-1:0] ddr_a,
output wire [((1<<DQ_LEVEL)+1)/2-1:0] ddr_dm,
inout [((1<<DQ_LEVEL)+1)/2-1:0] ddr_dqs,
inout [ (4<<DQ_LEVEL)-1:0] ddr_dqIt can be seen that the bit width of some signals of the DDR1 interface is related to the parameters, and the user needs to determine the module parameters according to the DDR1 chip part. For details, see [Bit Width Parameters](#Bit Width Parameters).
If you want to know the waveforms of the DDR1 interface during initialization, read/write, and refresh, please perform [RTL Simulation](#RTL Simulation).
The DDR1 controller provides clock and reset signals for the AXI4 slave interface, as follows. The AXI4 masters should use them as their clock and reset.
output reg rstn,
output reg clk,Following code are the AXI4 slave interfaces of this module, they are all synchronized with the rising edge of clk clock and should be connected to the AXI4 master inside the FPGA.
input wire awvalid,
output wire awready,
input wire [BA_BITS+ROW_BITS+COL_BITS+DQ_LEVEL-2:0] awaddr,
input wire [ 7:0] awlen,
input wire wvalid,
output wire wready,
input wire wlast,
input wire [(8<<DQ_LEVEL)-1:0] wdata,
output wire bvalid,
input wire bready,
input wire arvalid,
output wire arready,
input wire [BA_BITS+ROW_BITS+COL_BITS+DQ_LEVEL-2:0] araddr,
input wire [ 7:0] arlen,
output wire rvalid,
input wire rready,
output wire rlast,
output wire [(8<<DQ_LEVEL)-1:0] rdata,The bit width of the data (wdata and rdata) of the AXI4 bus is twice that of the DDR1 data bit width. For example, for MT46V64M8 whose bit width is 8bit, the bit width of wdata and rdata is 16bit.
The addresses (awaddr and araddr) of the AXI4 are byte addresses. The module will calculate the bit width of awaddr and araddr according to the parameters specified by the user = BA_BITS+ROW_BITS+COL_BITS+DQ_LEVEL-1 . For example, for a 64MB MT46V64M8, the awaddr/araddr bit width is 26, where 2^26=64MB.
For each read and write access, the address value (awaddr and araddr) must be aligned with the entire word. For example, if the bit width of wdata and rdata is 16bit (2 bytes), then awaddr and araddr should be aligned by 2 bytes, that is, Divide by 2.
The AXI4 interface of this module does not support byte strobe write, there is no wstrb signal, and at least one entire data bit width is written each time. For example, if the bit width of wdata is 16, then at least 2 bytes are written at a time.
The AXI4 interface of this module supports burst reading and writing, and the burst length can take any value between 1~256, that is, the value range of awlen and arlen is 0~255. For example, in a write operation, awlen=12, the burst length is 13, and if the bit width of wdata is 16, then the burst write writes a total of 13*2=26B.
Note that each burst read and write cannot exceed a row boundary within DDR1. For example, for MT46V64M8, each row has 8*2^11/8 = 2048B, so each burst read and write cannot exceed the 2048B boundary.
__ __ __ __ __ __ __ __ __ __ __ __ _
clk __/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/
___________
awvalid ___/ \____________________________________________________________
_____
awready _________/ \____________________________________________________________
___________
awaddr XXXX____ADDR___XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
___________
awlen XXXX____8'd3___XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
___________ ___________ _____
wvalid _______________/ \_____/ \_____/ \__________________
___________________________________
wready _____________________/ \__________________
_____
wlast ___________________________________________________/ \__________________
___________ _____ _____ _____
wdata XXXXXXXXXXXXXXXX_____D0____XXXXXXX__D1_X__D2_XXXXXXX__D3_XXXXXXXXXXXXXXXXXXX
___________
bvalid _________________________________________________________/ \______
_____
bready _______________________________________________________________/ \______
一个典型的，突发长度为 4 （awlen=3） 的 AXI4 写操作如上图。分为 3 步：
- 地址通道握手：AXI4 主机把 awvalid 信号拉高，指示想要发起写动作，图中经过一周期后，DDR1 控制器才把 awready 信号拉高，说明地址通道握手成功（在此之前，DDR1 控制器可能在处理上一次读写动作，或者在进行刷新，因此暂时没把 awready 拉高）。握手成功的同时，DDR1 控制器收到待写地址（awaddr）和突发长度（awlen），awlen=8'd3 说明突发长度为 4。
- 数据传送：AXI4 总线上的 wvalid 和 wready 进行 4 次握手，把突发传输的 4 个数据写入 DDR1。在握手时，若 AXI4 主机还没准备好数据，可以把 wvalid 拉低，若 DDR1 控制器还没准备好接受数据，会把 wready 拉低。注意，根据 AXI4 总线的规定，在突发传输的最后一个数据传输的同时需要把 wlast 信号置高（实际上即使不置高，DDR1 控制器也能根据突发长度自动结束数据传送，进入写响应状态）。
- 写响应：数据传送结束后，DDR1 控制器还需要和 AXI4 主机进行一次握手，才能彻底结束本次写动作。DDR1 控制器在数据传送结束后立即把 wvalid 信号拉高，并等待 wready 为高。完成该握手后，DDR1才能响应下一次读写操作。
__ __ __ __ __ __ __ __ __ __ __ __ _
clk __/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/
___________
arvalid ___/ \____________________________________________________________
_____
arready _________/ \____________________________________________________________
___________
araddr XXXX____ADDR___XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
___________
arlen XXXX____8'd4___XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
___________________________________
rvalid _________________________________/ \______
___________________________________________________ __________________
rready \_____/
_____
rlast _______________________________________________________________/ \______
_____ _____ _____ ___________ _____
rdata XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX__D0_X__D1_X__D2_X_____D3____X__D4_XXXXXXX
A typical AXI4 read operation with a burst length of 5 (arlen=4) is shown above. Divided into 2 steps:
- Address channel handshake: The AXI4 host pulls the
arvalidsignal high, indicating that it wants to initiate a write action. After a cycle in the figure, the DDR1 controller pulls the ready signal high, indicating that the address channel handshake is successful (before this , the DDR1 controller may be processing the last read or write action, or performing a refresh, so it has not pulled ready for the time being). At the same time as the handshake is successful, the DDR1 controller receives the address to be read (araddr) and the burst length (arlen).arlen=8'd4means that the burst length is 5. - Data transfer:
rvalidandrreadyon the AXI4 bus perform 5 handshakes, and read out 5 of the burst transfer. During handshake, if the AXI4 host is not ready to accept data,rreadycan be pulled low, and the DDR1 controller will keep the current data until the AXI4 host can accept the data (that is, pullrreadyhigh). When the last data is transferred, the DDR1 controller will pullrlasthigh.
Note: When the module parameter
READ_BUFFER=0, the module will save a BRAM resource and also reduce the delay between address channel handshake and data transfer. But the DDR1 controller ignores therready=0condition and does not wait for the AXI4 host to be ready to accept data. This will destroy the completeness of the AXI4 protocol, but it may be useful in some simple situations.
This section describes how to determine the 4 parameters: BA_BITS , ROW_BITS , COL_BITS and DQ_LEVEL.
Take MICRON's DDR-SDRAM series chips as an example, different chips have different ROW ADDRESS BITS, COL ADDRESS BITS and DATA BITS, which means that their bit width parameters are also different, as shown in the following table (note: these parameters can be found in the chip datasheet).
| chip part | ddr_dq width | ddr_dm ddr_dqs width | DQ_LEVEL | BA_BITS | ROW_BITS | COL_BITS | byte per row | total capacity | awaddr/araddr width |
|---|---|---|---|---|---|---|---|---|---|
| MT46V64M4 | 4 | 1 | 0 | 2 | 13 | 11 | 1024 | 4*2^(2+13+11)=256Mb=32MB | 25 |
| MT46V128M4 | 4 | 1 | 0 | 2 | 13 | 12 | 2048 | 4*2^(2+13+12)=512Mb=64MB | 26 |
| MT46V256M4 | 4 | 1 | 0 | 2 | 14 | 12 | 4096 | 4*2^(2+14+12)=1Gb=128MB | 27 |
| MT46V32M8 | 8 | 1 | 1 | 2 | 13 | 10 | 1024 | 8*2^(2+13+10)=256Mb=32MB | 25 |
| MT46V64M8 | 8 | 1 | 1 | 2 | 13 | 11 | 2048 | 8*2^(2+13+11)=512Mb=64MB | 26 |
| MT46V128M8 | 8 | 1 | 1 | 2 | 14 | 11 | 4096 | 8*2^(2+14+11)=1Gb=128MB | 27 |
| MT46V16M16 | 16 | 2 | 2 | 2 | 13 | 9 | 1024 | 16*2^(2+13+9)=256Mb=32MB | 25 |
| MT46V32M16 | 16 | 2 | 2 | 2 | 13 | 10 | 2048 | 16*2^(2+13+10)=512Mb=64MB | 26 |
| MT46V64M16 | 16 | 2 | 2 | 2 | 14 | 10 | 4096 | 16*2^(2+14+10)=1Gb=128MB | 27 |
This section describes how to determine the 3 timing parameters tREFC, tW2I and tR2I.
We know that DDR1 requires periodic refresh action, and tREFC specifies the refresh clock cycle interval (subject to clk). For example, if the user clock is 75MHz, according to the chip manual of MT46V64M8, it needs to be refreshed once at most 7.8125us. Considering that 75MHz * 7.8125us = 585.9, this parameter can be set to a value less than 585, such as 10'd512.
tW2I specifies the minimum number of clock cycles (in clk) from the last write command of a write operation to the active command (ACT) of the next operation. The following figure shows a write operation on a DDR1 interface. The first rising edge of ddr_cas_n represents the end of the last write command of a write operation, and the second falling edge of ddr_ras_n represents the next operation (may be read, write, Refresh), there are 5 clock cycles between them, tW2I is used to specify the lower limit of the number of cycles. The default value of tW2I is 8'd7, which is a conservative value compatible with most DDR1s. For different DDR1 chips, there are different shrinkage margins (see DDR1 chip datasheet for details).
__ __ __ __ __ __ __ __ __ __ __ __
ddr_ck_p __/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__
_____ _______________________________________________ ________
ddr_ras_n \_____/ \_____/
_________________ ____________________________________________
ddr_cas_n \___________/
_________________ ____________________________________________
ddr_we_n \___________/
_____
ddr_a[10] XXXXXXXXXXXXXXXXXX_____/ XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
_______________________
ddr_ba XXXXXX__________BA___________XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
_______________________
ddr_a XXXXXX__RA_XXXXXXX_CA0_X_CA1_XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
tR2I specifies the minimum number of clock cycles (in clk) from the last read command of a read operation to the active command (ACT) of the next operation. The figure below shows a read operation on a DDR1 interface. The first rising edge of ddr_cas_n represents the end of the last read command of a read operation, and the second falling edge of ddr_ras_n represents the next operation (may be read, write, Refresh), there are 5 clock cycles between them, tR2I is used to specify the lower limit of the number of cycles. The default value of tR2I is 8'd7, which is a conservative value compatible with most DDR1. For different DDR1 chips, there are different shrinkage margins (see DDR1 chip datasheet for details).
__ __ __ __ __ __ __ __ __ __ __ __
ddr_ck_p __/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__
_____ _______________________________________________ ________
ddr_ras_n \_____/ \_____/
_________________ ____________________________________________
ddr_cas_n \___________/
__________________________________________________________________________
ddr_we_n
_____
ddr_a[10] XXXXXXXXXXXXXXXXXX_____/ XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
_______________________
ddr_ba XXXXXX__________BA___________XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
_______________________
ddr_a XXXXXX__RA_XXXXXXX_CA0_X_CA1_XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
I provide a DDR1 read and write self-test project based on the [FPGA+DDR1 board](#Hardware Design Example) I designed. The project directory is example-selftest , please open it with Quartus.
This project contains following files:
| File Name | Introduction |
|---|---|
| example-selftest/fpga_top.v | top-level module |
| example-selftest/axi_self_test_master.v | AXI4 master. First, write the data into DDR1 through AXI4, and then read back and verify. |
| RTL/ddr_sdram_ctrl.v | DDR1 controller |
The behavior of the demo project is:
Write: After the project starts running, it will first write the entire DDR1 through AXI4, and directly write the address word to the corresponding address. For example, if the data width of AXI4 is 16bit, then write 0x0001 at address 0x000002. Write 0x3456 at address 0x123456.
Read & Error Check: After writing the entire DDR1, the project will repeatedly read the entire DDR1 round by round. error) to generate a high-level pulse, and the error count signal (error_cnt) +1. If the DDR1 configuration is correct, there should be no error signal. You can measure the pin corresponding to error_cnt, if it is 0 (all low), it means there is no error.
SignalTap waveform capture: This project contains a SignalTap file stp1.stp, which can be used to view the waveform on the DDR1 interface when the program is running. It triggers with error=1, so if there is no error in the read and write self-test, it will not be triggered. Because the project is reading DDR1 at any time, if you want to see the waveform on the DDR1 interface, just press the "Stop" button.
Modify AXI4 burst length: You can modify WBURST_LEN and RBURST_LEN in lines 82 and 83 of fpga_top.v to modify the write/read burst length during self-test. The self-test program only supports 2^n-1 This burst length, ie WBURST_LEN and RBURST_LEN must take values like 0, 1, 3, 7, 15, 31, ... (note that this is only a limitation of the self-test program I wrote, DDR1 controller supports 0 Any burst length between ~255.
WBURST_LEN and RBURST_LEN can be set differently.
I provide a UART read and write project based on the [FPGA+DDR1 board](#Hardware Design Example) I designed. In this project, you can read and write DDR1 with different burst lengths through UART commands. The project directory is example-uart-read-write , please open it with Quartus.
The project contains the following files:
| File Name | Introduction |
|---|---|
| example-uart-read-write/fpga_top.v | top-level module |
| example-uart-read-write/uart/uart2axi4.v | an AXI4 master, which can convert the commands received by UART RX into AXI4 read and write operations, and send the data read out by read operations through UART TX. see https://github.com/WangXuan95/Verilog-UART for detail. |
| example-uart-read-write/uart/uart_rx.v | called by uart2axi4.v |
| example-uart-read-write/uart/uart_tx.v | called by uart2axi4.v |
| RTL/ddr_sdram_ctrl.v | DDR1 controller |
There is a CH340E chip (USB to UART) on the [FPGA+DDR1 board](#Hardware Design Example), so after plugging in the USB cable, you can see the serial port corresponding to the UART on the computer (you need to download and install the driver of CH341 on www.wch. cn/product/CH340.html first).
After the project is programed to FPGA, double-click to open a serial port tool UartSession.exe (it is in the example-uart-read-write directory), open the COM port corresponding to the board according to the prompt, and then type the following command + Enter, you can Write the 4 data 0x0123, 0x4567, 0x89ab, and 0xcdef into the starting address 0x12345. (A write operation with a burst length of 4 occurs on the AXI4 bus).
w12345 0123 4567 89ab cdef
Then use the following command + enter to read 8 data from the starting address 0x12345.
r12345 8
The effect is as shown in the figure below. The first 4 data (0123 4567 89ab cdef) are what we have written into DDR1, and the last 4 data are random data that comes with DDR1 after initialization.
 |
|---|
| Figure : Use UartSession.exe to perform DDR1 read/write. |
The length of the write burst is how much data there is in the write command. For example, the burst length of the following write command is 10, and 10 pieces of data are written to the starting address 0x00000
w0 1234 2345 3456 4567 5678 6789 789a 89ab 9abc abcd
The read command needs to specify the burst length. For example, the burst length of the following command is 30 (0x1e), and 30 data are read from the starting address 0x00000
r0 1e
The files required for the simulation are in the SIM folder, where:
| File Name | Introduction |
|---|---|
| tb_ddr_sdram_ctrl.v | Testbench code, simulation top-level. |
| axi_self_test_master.v | AXI4 master. First, write the data into DDR1 through AXI4, and then read back and verify. |
| micron_ddr_sdram_model.v | MICRON's DDR1 simulation model |
The behavior of the simulation project is the same as that of the self-test program, axi_self_test_master.v, as the AXI4 host, writes regular data into DDR1, but not all of them, only the first 16KB of DDR1 (because the memory spave of the simulation model is limited), and then repeatedly read data round by round to compare whether there is unmatched data, if there is, generate a high level of one clock cycle on the error signal.
Before using iverilog for simulation, you need to install iverilog , see: iverilog_usage
Then double-click tb_ddr_sdram_ctrl_run_iverilog.bat to run the simulation, and then you can open the generated dump.vcd file to view the waveform.
The default configuration parameters of the above simulation fit MT46V64M8, namely ROW_BITS=13, COL_BITS=11, and DQ_BITS=8. If you want to simulate other DDR1 chips, you need to modify them in tb_ddr_sdram_ctrl.v. For MICRON's DDR1 series, these parameters should be modified as follows:
| DDR1 part | BA_BITS | ROW_BITS | COL_BITS | DQ_LEVEL |
|---|---|---|---|---|
| MT46V64M4 | 2 | 13 | 11 | 0 |
| MT46V128M4 | 2 | 13 | 12 | 0 |
| MT46V256M4 | 2 | 14 | 12 | 0 |
| MT46V32M8 | 2 | 13 | 10 | 1 |
| MT46V64M8 | 2 | 13 | 11 | 1 |
| MT46V128M8 | 2 | 14 | 11 | 1 |
| MT46V16M16 | 2 | 13 | 9 | 2 |
| MT46V32M16 | 2 | 13 | 10 | 2 |
| MT46V64M16 | 2 | 14 | 10 | 2 |
In addition, you can modify lines 18~19 of tb_ddr_sdram_ctrl.v to modify the burst length of reads and writes during simulation.
- MICRON's DDR1 simulation model : https://www.micron.com/products/dram/ddr-sdram/part-catalog/mt46v64m8p-5b
- MT46V64M8 datasheet : https://media-www.micron.com/-/media/client/global/documents/products/data-sheet/dram/ddr1/512mb_ddr.pdf?rev=4e1e995d6d2240e293286770f193d57d
DDR1 SDRAM 控制器，具有 AXI4 slave 接口，接收 AXI4 读写操作进行 DDR1 读写。
在低端FPGA设计中用 DDR-SDRAM（DDR1）替换 SDR-SDRAM。
|------------------------------| |-----------|
| | | |
----->| driving clock and reset | | |
| | | |
------------| | DDR-SDRAM interface |--------->| |
| AXI4 | | | |
AXI4 master |-------->| AXI4 slave | | |
| | | | |
------------| |------------------------------| |-----------|
用户逻辑 DDR1 控制器 DDR1 芯片
很多低端 FPGA 开发板使用 SDR-SDRAM 作为片外存储，而 DDR-SDRAM (DDR1) 比 SDR-SDRAM 容量更大，价格更低。且与SDR-SDRAM一样，DDR1也能使用低端FPGA的普通的IO管脚直接驱动。我编写了一个软核的 AXI4 接口的 DDR1 控制器。该控制器的特点有：
- 平台无关 ：纯 Verilog 编写，可以在 Altera 和 Xilinx 等各种 FPGA 上运行。
- 兼容性强 ：支持各种位宽和容量的 DDR1 （已在MICRON所有位宽和容量的DDR1型号上仿真通过）。
为了展示该控制器的使用方法，我提供了两个示例程序：
- 自测程序：通过DDR1控制器，将数据写入 DDR1 ，然后读出，比较读出的数据是否与写入的吻合。
- UART读写程序：将 UART 命令转化成 AXI4 总线动作，通过DDR1控制器读写DDR1。
另外，由于各代 DDR-SDRAM（例如DDR3、DDR2、DDR1）的接口时序大同小异，本库也可以方便那些熟悉 Verilog 的人来学习 DDR-SDRAM 接口。
对于 FPGA 的选型，只需要有足够数量的普通 IO 的 FPGA 就可以驱动 DDR1。DDR1 的 IO 电平标准往往是 SSTL-2 ，与 2.5V LVTTL 或 2.5V LVCMOS 兼容，因此相应的 FPGA 的 IO bank 的电源应该是 2.5V，且应在开发软件配置为 2.5V LVTTL 或 2.5V LVCMOS 。
下表展示了 DDR1 颗粒 的引脚，以及其与 FPGA 连接时应该注意的点。
| 引脚名称 | 方向 | 位宽 | 说明 | 电平 | 注意事项 |
|---|---|---|---|---|---|
| ddr_ck_p | FPGA output | 1 | 差分时钟正，≥75MHz | 2.5V LVTTL | 布线尽量短，与 ddr_ck_n 差分布线 |
| ddr_ck_n | FPGA output | 1 | 差分时钟负，≥75MHz | 2.5V LVTTL | 布线尽量短，与 ddr_ck_p 差分布线 |
| ddr_cke、ddr_cs_n | FPGA output | 1 | 低速 | 2.5V LVTTL | |
| ddr_ras_n、ddr_cas_n、ddr_we_n | FPGA output | 1 | 与 ddr_ck_p 下降沿同步 | 2.5V LVTTL | 布线尽量短，与 ddr_ck_p 大致等长（长度别太离谱即可） |
| ddr_ba | FPGA output | 2 | 与 ddr_ck_p 下降沿同步 | 2.5V LVTTL | 布线尽量短，与 ddr_ck_p 大致等长（别太离谱即可） |
| ddr_a | FPGA output | 取决于芯片型号 | 与 ddr_ck_p 下降沿同步 | 2.5V LVTTL | 布线尽量短，与 ddr_ck_p 大致等长（别太离谱即可） |
| ddr_dqs | inout | 取决于芯片型号 | 2.5V LVTTL | 布线尽量短 | |
| ddr_dm、ddr_dq | inout | 取决于芯片型号 | 与 ddr_dqs 双沿同步 | 2.5V LVTTL | 布线尽量短，与 ddr_dqs 大致等长（别太离谱即可） |
为了进行展示，我用 Altera Cyclone IV 的最低廉的 FPGA （型号：EP4CE6E22C8N) 和 MICRON 的 64MB DDR1 （型号 MT46V64M8TG-6T）画了一个小板子，本库的所有例子可以直接在该板子上直接运行。（如果你要在自己的PCB设计中使用 DDR1 ，只要抄我这个板子就行）
|
|---|
| 图：FPGA + DDR1 测试板 |
该板子的原理图见 PCB/sch.pdf ，PCB制造文件见 PCB/gerber.zip 。它是个双层板，不需像 DDR2 、 DDR3 那样刻意注意等长和阻抗匹配，因为该电路工作频率为双边沿 75MHz，不是特别高，只需注意让FPGA与DDR距离尽量近，布线尽量短即可。比如我把DDR1芯片放在了FPGA芯片正对的背面，从而保证布线都较短。
该板子的设计在立创EDA中开放，见 oshwhub.com/wangxuan/fpga-ddr-ce-shi-ban。
DDR1 控制器代码见 RTL/ddr_sdram_ctrl.v ，它能自动对 DDR1 进行初始化，并定时进行刷新（Refresh）。该模块有一个简化但完备的 AXI4 从接口，通过它可以完成对 DDR1 的读写。本节详细解释该模块的使用方法。
模块参数和默认值在 Verilog 中定义如下：
module ddr_sdram_ctrl #(
parameter READ_BUFFER = 1,
parameter BA_BITS = 2,
parameter ROW_BITS = 13,
parameter COL_BITS = 11,
parameter DQ_LEVEL = 1,
parameter [9:0] tREFC = 10'd256,
parameter [7:0] tW2I = 8'd7,
parameter [7:0] tR2I = 8'd7
)对这些参数说明如下：
| 参数名 | 类型 | 取值范围 | 默认值 | 说明 |
|---|---|---|---|---|
| READ_BUFFER | 配置参数 | 0 或 1 | 1 | 若设为0，DDR1控制器内将没有读数据缓冲，读出的数据将不会等待AXI4主机是否能接受，即rvalid信号不会等待rready信号，直接以最高速率灌出去，此时AXI4是不完备的，但可以降低读出的时延。若设为1，则DDR1控制器内会有一个足够大的读数据缓冲，rvalid信号会与rready信号握手，确认AXI4主机准备好后再读出数据。 |
| BA_BITS | 位宽参数 | 1~4 | 2 | 规定了 DDR BANK ADDRESS (ddr_ba) 的宽度，常规的 DDR1 的 BANK ADDRESS 都是 2bit，因此该参数通常固定为默认值不用改 |
| ROW_BITS | 位宽参数 | 1~15 | 13 | 规定了 DDR ROW ADDRESS 的宽度，同时也决定了 DDR1 芯片的地址线引脚(ddr_a)的宽度。该参数取决于DDR1芯片的选型。例如 MT46V64M8 每个 BANK 有 8192 个 COL，考虑到2^11=8192，该参数应为13。同理，对于 MT46V128M8，该参数应为14 |
| COL_BITS | 位宽参数 | 1~14 | 11 | 规定了 DDR COL ADDRESS 的宽度。该参数取决于DDR1芯片的选型。例如 MT46V64M8 每个 ROW 有 2048 个 COL，考虑到2^11=2048，因此该参数应该为11。同理，对于MT46V32M16，该参数应为10。 |
| DQ_LEVEL | 位宽参数 | 0~7 | 1 | 规定了 DDR1 芯片的数据位宽，对于位宽x4的芯片（例如MT46V64M4），该参数应取0；对于位宽x8的芯片（例如MT46V64M8），该参数应取1；对于位宽x16的芯片（例如MT46V64M16），该参数应取2；对于位宽x32的情况（例如两片MT46V64M16扩展位宽），该参数应取3；以此类推。 |
| tREFC | 时序参数 | 10'd1 ~ 10'd1023 | 10'd256 | 该控制器会周期性地刷新 DDR1，该参数规定了刷新的间隔。详见下文时序参数的确定 |
| tW2I | 时序参数 | 8'd1~8'd255 | 8'd7 | 该参数规定了一个写动作的最后一个写命令到下一个动作的激活命令(ACT)的间隔。详见下文时序参数的确定 |
| tR2I | 时序参数 | 8'd1~8'd255 | 8'd7 | 该参数规定了一个读动作的最后一个读命令到下一个动作的激活命令(ACT)的间隔。详见下文时序参数的确定 |
该模块需要一个驱动时钟和一个驱动复位，如下：
input wire rstn_async,
input wire drv_clk,rstn_async 是低电平复位信号，正常工作时应该置高。drv_clk 是驱动时钟，频率是用户时钟的 4 倍。
模块开始工作前，rstn_async 信号应该置低，让模块复位，然后把 rstn_async 置高，解除复位。
模块内将驱动时钟 drv_clk 分频 4 倍产生 DDR1 时钟（ddr_ck_p/ddr_ck_n）和 AXI4 总线用户时钟（clk）。本节讲述如何决定驱动时钟 drv_clk 的频率。
首先，时钟频率受限于 DDR1 芯片。考虑到所有的 DDR1 的接口频率至少为 75MHz ，则 drv_clk 的下限是 75*4=300MHz。
而 drv_clk 的上限就也取决于 DDR1 的芯片型号，例如对于 MT46V64M8P-5B ，查芯片手册可知，-5B 后缀的 DDR1 在 CAS Latency (CL)=2 时最高时钟频率是 133MHz，则 drv_clk 的上限是 133*4=532MHz 。
注意：本控制器固定 CAS Latency (CL) = 2。
另外，时钟频率的上限还受限于 FPGA 的速度，太高的时钟频率容易导致时序不收敛。本设计充分考虑时序安全设计，大多数寄存器工作在频率较低 clk 时钟域；个别寄存器工作在 clk 时钟的 2 倍频率的时钟下，且输入端口的组合逻辑非常短；还有一个寄存器工作在高频的 drv_clk 下，但输入端口直接来自于上一级的寄存器输出（没有组合逻辑）。因此，即使在速度级别很低的 EP4CE6E22C8N 上，在 300MHz 的驱动时钟下也能保证模块正确运行。在速度等级更高的 FPGA （例如 EP4CE22F17C6N）上，驱动时钟的频率可以更高（例如400MHz）。
以下是该模块的 DDR1 接口。这些接口应该直接从 FPGA 引出，连接到 DDR1 芯片上。
output wire ddr_ck_p, ddr_ck_n,
output wire ddr_cke,
output reg ddr_cs_n,
output reg ddr_ras_n,
output reg ddr_cas_n,
output reg ddr_we_n,
output reg [ BA_BITS-1:0] ddr_ba,
output reg [ ROW_BITS-1:0] ddr_a,
output wire [((1<<DQ_LEVEL)+1)/2-1:0] ddr_dm,
inout [((1<<DQ_LEVEL)+1)/2-1:0] ddr_dqs,
inout [ (4<<DQ_LEVEL)-1:0] ddr_dq可以看出 DDR1 接口的一些信号的位宽是和参数有关的，用户需要根据 DDR1 的芯片选型来确定模块参数。详见 位宽参数的确定。
想了解 DDR1 接口在初始化、读写、刷新时的波形，请进行 仿真。
DDR1 控制器对外提供 AXI4 从接口（AXI4 slave）的时钟和复位信号，如下。AXI4 主机应该用它们作为自己的时钟和复位。
output reg rstn,
output reg clk,以下是该模块的 AXI4 从接口，它们都与 clk 时钟的上升沿同步，应该连接到 FPGA 内的的 AXI4 主机。
input wire awvalid,
output wire awready,
input wire [BA_BITS+ROW_BITS+COL_BITS+DQ_LEVEL-2:0] awaddr,
input wire [ 7:0] awlen,
input wire wvalid,
output wire wready,
input wire wlast,
input wire [(8<<DQ_LEVEL)-1:0] wdata,
output wire bvalid,
input wire bready,
input wire arvalid,
output wire arready,
input wire [BA_BITS+ROW_BITS+COL_BITS+DQ_LEVEL-2:0] araddr,
input wire [ 7:0] arlen,
output wire rvalid,
input wire rready,
output wire rlast,
output wire [(8<<DQ_LEVEL)-1:0] rdata,AXI4 总线的数据（wdata和rdata）的位宽是 DDR1 数据位宽的 2 倍。例如，对于位宽为 8bit 的 MT46V64M8，wdata 和 rdata 的位宽是 16bit 。
AXI4 总线的地址（awaddr和araddr）统一是字节地址，模块会根据用户指定的参数来计算 awaddr 和 araddr 的位宽 = BA_BITS+ROW_BITS+COL_BITS+DQ_LEVEL-1 。例如对于 64MB 的 MT46V64M8 ，awaddr/araddr 位宽为 26，而 2^26 等于 64MB。
每次读写访问时，地址值（awaddr和araddr）都要和整个字对齐，例如若 wdata 和 rdata 的位宽是 16bit（2字节），那么 awaddr 和 araddr 要按 2 字节对齐，即能整除 2。
该模块的 AXI4 接口不支持字节选通写入，没有 wstrb 信号，每次至少写入一整个数据位宽，例如若 wdata 的位宽是 16，那么一次至少写入2字节。
该模块的 AXI4 接口支持突发读写，突发长度可取 1~256 之间的任意值，即 awlen 和 arlen 取值范围为 0~255 。例如，在一次写入动作中，awlen=12，则突发长度为 13，若 wdata 的位宽是 16 ，那么该次突发写一共写了 13*2=26B。
另外，每次突发读写不能超越 DDR1 内的一个行的边界。例如对于 MT46V64M8，每个行有 8*2^11/8 = 2048B，那么每次突发读写就不能超越 2048B 的边界。
__ __ __ __ __ __ __ __ __ __ __ __ _
clk __/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/
___________
awvalid ___/ \____________________________________________________________
_____
awready _________/ \____________________________________________________________
___________
awaddr XXXX____ADDR___XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
___________
awlen XXXX____8'd3___XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
___________ ___________ _____
wvalid _______________/ \_____/ \_____/ \__________________
___________________________________
wready _____________________/ \__________________
_____
wlast ___________________________________________________/ \__________________
___________ _____ _____ _____
wdata XXXXXXXXXXXXXXXX_____D0____XXXXXXX__D1_X__D2_XXXXXXX__D3_XXXXXXXXXXXXXXXXXXX
___________
bvalid _________________________________________________________/ \______
_____
bready _______________________________________________________________/ \______
一个典型的，突发长度为 4 （awlen=3） 的 AXI4 写操作如上图。分为 3 步：
- 地址通道握手：AXI4 主机把 awvalid 信号拉高，指示想要发起写动作，图中经过一周期后，DDR1 控制器才把 awready 信号拉高，说明地址通道握手成功（在此之前，DDR1 控制器可能在处理上一次读写动作，或者在进行刷新，因此暂时没把 awready 拉高）。握手成功的同时，DDR1 控制器收到待写地址（awaddr）和突发长度（awlen），awlen=8'd3 说明突发长度为 4。
- 数据传送：AXI4 总线上的 wvalid 和 wready 进行 4 次握手，把突发传输的 4 个数据写入 DDR1。在握手时，若 AXI4 主机还没准备好数据，可以把 wvalid 拉低，若 DDR1 控制器还没准备好接受数据，会把 wready 拉低。注意，根据 AXI4 总线的规定，在突发传输的最后一个数据传输的同时需要把 wlast 信号置高（实际上即使不置高，DDR1 控制器也能根据突发长度自动结束数据传送，进入写响应状态）。
- 写响应：数据传送结束后，DDR1 控制器还需要和 AXI4 主机进行一次握手，才能彻底结束本次写动作。DDR1 控制器在数据传送结束后立即把 wvalid 信号拉高，并等待 wready 为高。完成该握手后，DDR1才能响应下一次读写操作。
__ __ __ __ __ __ __ __ __ __ __ __ _
clk __/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/
___________
arvalid ___/ \____________________________________________________________
_____
arready _________/ \____________________________________________________________
___________
araddr XXXX____ADDR___XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
___________
arlen XXXX____8'd4___XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
___________________________________
rvalid _________________________________/ \______
___________________________________________________ __________________
rready \_____/
_____
rlast _______________________________________________________________/ \______
_____ _____ _____ ___________ _____
rdata XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX__D0_X__D1_X__D2_X_____D3____X__D4_XXXXXXX
一个典型的，突发长度为 5 （arlen=4） 的 AXI4 读操作如上图。分为 2 步：
- 地址通道握手：AXI4 主机把 arvalid 信号拉高，指示想要发起写动作，图中经过一周期后，DDR1 控制器才把 arready 信号拉高，说明地址通道握手成功（在此之前，DDR1 控制器可能在处理上一次读写动作，或者在进行刷新，因此暂时没把 arready 拉高）。握手成功的同时，DDR1 控制器收到待读地址（araddr）和突发长度（arlen），arlen=8'd4 说明突发长度为 5。
- 数据传送：AXI4 总线上的 rvalid 和 rready 进行 5 次握手，把突发传输的 5 个读出。在握手时，若 AXI4 主机还没准备好接受数据，可以把 rready 拉低，则 DDR1 控制器会保持当前数据，直到 AXI4主机能接受此数据（即把rready拉高）为止。在传送最后一个数据时，DDR1 控制器会把 rlast 拉高。
注：当模块参数 READ_BUFFER = 0 时，模块将节省一个 BRAM 资源，同时还能降低地址通道握手和数据传送之间的延迟。但 DDR1 控制器会忽略 rready=0 的情况，不等待 AXI4 主机是否准备好接受数据。这样将会破坏 AXI4 协议的完备性，但在一些简单的场合或许有用。
本节讲述如何确定 BA_BITS、ROW_BITS、COL_BITS 和 DQ_LEVEL 这 4 个参数。
从模块接口：DDR1接口一节可以看出 DDR1 接口的一些信号的位宽是和参数有关的，用户需要根据 DDR1 的芯片选型来确定模块参数。
以 MICRON 公司的 DDR-SDRAM 系列芯片为例，不同芯片有不同的 ROW ADDRESS BITS (行地址宽度）、COL ADDRESS BITS（列地址宽度）、DATA BITS（数据宽度），决定了他们的位宽参数也不同，如下表（注：这些参数都能从芯片datasheet中查到）。
| 芯片名称 | ddr_dq 位宽 | ddr_dm 和 ddr_dqs 位宽 | DQ_LEVEL 取值 | BA_BITS | ROW_BITS | COL_BITS | 每行字节数 | 容量 | awaddr/araddr 位宽 |
|---|---|---|---|---|---|---|---|---|---|
| MT46V64M4 | 4 | 1 | 0 | 2 | 13 | 11 | 1024 | 4*2^(2+13+11)=256Mb=32MB | 25 |
| MT46V128M4 | 4 | 1 | 0 | 2 | 13 | 12 | 2048 | 4*2^(2+13+12)=512Mb=64MB | 26 |
| MT46V256M4 | 4 | 1 | 0 | 2 | 14 | 12 | 4096 | 4*2^(2+14+12)=1Gb=128MB | 27 |
| MT46V32M8 | 8 | 1 | 1 | 2 | 13 | 10 | 1024 | 8*2^(2+13+10)=256Mb=32MB | 25 |
| MT46V64M8 | 8 | 1 | 1 | 2 | 13 | 11 | 2048 | 8*2^(2+13+11)=512Mb=64MB | 26 |
| MT46V128M8 | 8 | 1 | 1 | 2 | 14 | 11 | 4096 | 8*2^(2+14+11)=1Gb=128MB | 27 |
| MT46V16M16 | 16 | 2 | 2 | 2 | 13 | 9 | 1024 | 16*2^(2+13+9)=256Mb=32MB | 25 |
| MT46V32M16 | 16 | 2 | 2 | 2 | 13 | 10 | 2048 | 16*2^(2+13+10)=512Mb=64MB | 26 |
| MT46V64M16 | 16 | 2 | 2 | 2 | 14 | 10 | 4096 | 16*2^(2+14+10)=1Gb=128MB | 27 |
本节讲述如何确定 tREFC、tW2I 和 tR2I 这 3 个时序参数。
我们知道，DDR1 需要周期性的刷新动作，tREFC 就规定了刷新的时钟周期间隔（以 clk 为准）。例如，若用户时钟为 75MHz，根据 MT46V64M8 的芯片手册，需要至多 7.8125us 刷新一次，考虑到 75MHz * 7.8125us = 585.9，则该参数可以设置为一个小于 585 的值，例如 10'd512。
tW2I 规定了一个写操作的最后一个写命令到下一个操作的激活命令(ACT)的时钟周期数（以 clk 为准）的最小值。下图展示了一个 DDR1 接口上的写操作，ddr_cas_n 的第一个上升沿代表了一个写操作的最后一个写命令结束，ddr_ras_n 的第二个下降沿代表了下一个操作（可能是读、写、刷新）的开始，它们之间有 5 个时钟周期，tW2I 就是用来规定该周期数的下限的。tW2I 的默认值 8'd7 是一个兼容绝大多数 DDR1 的保守值，对于不同的 DDR1 芯片，有不同的缩小的余地（详见 DDR1 芯片 datasheet）。
__ __ __ __ __ __ __ __ __ __ __ __
ddr_ck_p __/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__
_____ _______________________________________________ ________
ddr_ras_n \_____/ \_____/
_________________ ____________________________________________
ddr_cas_n \___________/
_________________ ____________________________________________
ddr_we_n \___________/
_____
ddr_a[10] XXXXXXXXXXXXXXXXXX_____/ XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
_______________________
ddr_ba XXXXXX__________BA___________XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
_______________________
ddr_a XXXXXX__RA_XXXXXXX_CA0_X_CA1_XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
tR2I 规定了一个读操作的最后一个读命令到下一个操作的激活命令(ACT)的时钟周期数（以 clk 为准）的最小值。下图展示了一个 DDR1 接口上的读操作，ddr_cas_n 的第一个上升沿代表了一个读操作的最后一个读命令结束，ddr_ras_n 的第二个下降沿代表了下一个操作（可能是读、写、刷新）的开始，它们之间有 5 个时钟周期，tR2I 就是用来规定该周期数的下限的。tR2I 的默认值 8'd7 是一个兼容绝大多数 DDR1 的保守值，对于不同的 DDR1 芯片，有不同的缩小的余地（详见 DDR1 芯片 datasheet）。
__ __ __ __ __ __ __ __ __ __ __ __
ddr_ck_p __/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__/ \__
_____ _______________________________________________ ________
ddr_ras_n \_____/ \_____/
_________________ ____________________________________________
ddr_cas_n \___________/
__________________________________________________________________________
ddr_we_n
_____
ddr_a[10] XXXXXXXXXXXXXXXXXX_____/ XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
_______________________
ddr_ba XXXXXX__________BA___________XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
_______________________
ddr_a XXXXXX__RA_XXXXXXX_CA0_X_CA1_XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
我基于我画的 FPGA+DDR1测试板 做了一个 DDR1 读写自测程序，工程目录是 example-selftest ，请用 Quartus 打开它。
该工程包含以下文件：
| 文件名称 | 用途 |
|---|---|
| example-selftest/fpga_top.v | 顶层 |
| example-selftest/axi_self_test_master.v | 是 AXI4 主机，通过 AXI4 先把有规律的数据写入 DDR1，然后读回，比较读回的数据是否符合规律，并对不匹配的情况进行计数。 |
| RTL/ddr_sdram_ctrl.v | DDR1 控制器 |
该示例程序的行为是：
写入：该工程开始运行后，会先通过 AXI4 把整个 DDR1 都写一遍，直接把地址字写入相应的地址。例如，如果 AXI4 的数据宽度是 16bit，那么地址 0x000002 处写 0x0001。地址 0x123456 处写 0x3456。
读取&错误校验：写完整个 DDR1 后，该工程会一轮一轮地反复读取整个 DDR1，若读出的数据不符合上述规律，就认为出现错误，在错误信号（error）上产生一个高电平脉冲，并把错误计数信号（error_cnt）+1。如果 DDR1 配置正确，是不该出现 error 信号的。你可以测量 error_cnt 对应的引脚，若为0（全低电平），说明不存在错误。
SignalTap抓波形：该工程包含一个 SignalTap 文件 stp1.stp，在程序运行时，可以用它查看 DDR1 接口上的波形。它以 error=1 为触发信号，因此如果读写自测没有出错，它就不会被触发。因为该工程随时都在读取 DDR1，要想看 DDR1 接口上的波形，直接按“停止”按钮即可。
修改 AXI4 突发长度：在 fpga_top.v 的第 82，83 行可以修改 WBURST_LEN 和 RBURST_LEN，从而修改自测时的写/读突发长度，该自测程序只支持 2^n-1 这种突发长度，即 WBURST_LEN 和 RBURST_LEN 必须取形如 0,1,3,7,15,31,…… 的值（注意，这只是我编写的自测程序的限制，DDR1 控制器是支持 0~255 之间的任意突发长度的。
WBURST_LEN 和 RBURST_LEN 可以设置的不一样。
我基于我画的 FPGA+DDR1测试板 做了一个 UART 读写程序，使用该程序，你可以通过 UART 命令，以不同的突发长度来读写 DDR1。工程目录是 example-uart-read-write ，请用 Quartus 打开它。
该工程包含以下文件：
| 文件名称 | 用途 |
|---|---|
| example-uart-read-write/fpga_top.v | 顶层 |
| example-uart-read-write/uart/uart2axi4.v | 是 AXI4 主机，能把 UART RX 收到的命令转换成 AXI4 读写操作，并把读操作读出的数据通过 UART TX 发送出去 (详见 https://github.com/WangXuan95/Verilog-UART) |
| example-uart-read-write/uart/uart_rx.v | 被 uart2axi4.v 调用 |
| example-uart-read-write/uart/uart_tx.v | 被 uart2axi4.v 调用 |
| RTL/ddr_sdram_ctrl.v | DDR1 控制器 |
FPGA+DDR1测试板上有一个 CH340E 芯片（USB 转 UART），因此插上 USB 线后就可以在电脑上看见 UART 对应的 COM 口（需要先在 www.wch.cn/product/CH340.html 下载安装 CH341 的驱动）。
工程上传 FPGA 后，双击打开我编写的一个串口小工具 UartSession.exe （它在 example-uart-read-write 目录里），根据提示打开板子对应的 COM 口，然后打如下的命令+回车，可以把 0x0123 0x4567 0x89ab 0xcdef 这 4 个数据写入起始地址 0x12345。（AXI4总线上会产生一个突发长度为 4 的写操作）。
w12345 0123 4567 89ab cdef
然后用以下命令+回车，可以以 0x12345 为起始地址，以 8 为突发长度，读取 8 个数据。
r12345 8
效果如下图，前4个数据 (0123 4567 89ab cdef) 就是我们已经写入 DDR1 的，后4个数据我们没写过，是 DDR1 初始化后自带的随机数据。
|
|---|
| 图：使用 UartSession.exe 测试 DDR1 读写 |
写命令里有多少数据，写突发长度就是多少，例如以下写命令的突发长度是 10，将 10 个数据写入起始地址 0x00000
w0 1234 2345 3456 4567 5678 6789 789a 89ab 9abc abcd
读命令则直接指定突发长度，例如以下命令的突发长度为 30 （0x1e），从起始地址 0x00000 将 30 个数据读出
r0 1e
仿真所需要的文件在目录 SIM 里，其中：
| 文件路径 | 用途 |
|---|---|
| tb_ddr_sdram_ctrl.v | 仿真顶层 |
| axi_self_test_master.v | 是 AXI4 主机，通过 AXI4 先把有规律的数据写入 DDR1，然后读回，比较读回的数据是否符合规律，并对不匹配的情况进行计数。 |
| micron_ddr_sdram_model.v | MICRON 公司提供的 DDR1 仿真模型 |
该仿真工程的行为和自测程序一样， axi_self_test_master.v 作为 AXI4 主机，将有规律的数据写入 DDR1 中，只不过不是全部写入，而是只写入 DDR1 的前 16KB （因为仿真模型的存储空间有限），然后一轮一轮地反复读出数据，比较是否有不匹配的数据，若有，则在 error 信号上产生一个时钟周期的高电平。
使用 iverilog 进行仿真前，需要安装 iverilog ，见：iverilog_usage
然后双击 tb_ddr_sdram_ctrl_run_iverilog.bat 运行仿真，然后可以打开生成的 dump.vcd 文件查看波形。
以上仿真默认配置的参数是使用 MT46V64M8 ，即 ROW_BITS=13，COL_BITS=11，DQ_BITS=8 。如果想对其它型号的 DDR1 芯片进行仿真，你需要在 tb_ddr_sdram_ctrl.v 里修改它们。对于 MICRON 公司的 DDR1 系列，这些参数应该这样修改：
| 芯片名称 | BA_BITS | ROW_BITS | COL_BITS | DQ_LEVEL |
|---|---|---|---|---|
| MT46V64M4 | 2 | 13 | 11 | 0 |
| MT46V128M4 | 2 | 13 | 12 | 0 |
| MT46V256M4 | 2 | 14 | 12 | 0 |
| MT46V32M8 | 2 | 13 | 10 | 1 |
| MT46V64M8 | 2 | 13 | 11 | 1 |
| MT46V128M8 | 2 | 14 | 11 | 1 |
| MT46V16M16 | 2 | 13 | 9 | 2 |
| MT46V32M16 | 2 | 13 | 10 | 2 |
| MT46V64M16 | 2 | 14 | 10 | 2 |
另外，你可以修改 tb_ddr_sdram_ctrl.v 的第 18 和 19 行来修改仿真时的突发读写的长度。