Power-on and power-supply of DSP6000

1. DSP6000 power-on sequence Problems

DSP is indeed a freak. Two common problems are: (1) unable to connect to the CCS development environment; (2) burning chips. I am using TMS320C6713 now. The following experiences are based on DSP6000. Other series may be different. It seems that the performance of the chip has changed, and the chip has never been burned. However, CCS still often cannot be connected.

Some of the reasons are related to the stability of the hardware. When designing the power supply circuit, it is best to ensure that the kernel voltage of 1.2V is powered on first, and the IO port voltage of 3.3V is powered on later.

Another important reason is the DSP board. The connection sequence between DSP power switch and CCS is also very important. Previous connection Errors often lead to the inability to connect to CCS, CCS crashes and other inexplicable situations.

Summarize the correct power-on connection practices through an attempt:

(1) disconnect the DSP board, XDS510 simulator, and PC, and disable the CCS software.

(2) connect the XDS510 simulator to the DSP Development Board

(3) connect the XDS510 simulator to the USB port of the PC. Wait for a while (it is very important that there is an indicator on the XDS510 simulator, and wait until the indicator turns off)

(4) Open the CCS Software

(5) turn on the power switch of the DSP board and wait for a while (several seconds)

(6) Use Alt + C or Debug/Connect to Connect the simulator from CCS

In the above six steps, the connection success rate is still relatively high.

The power-off operation is also very important. Some people have not disconnected the CCS connection, and directly disconnected the simulator or DSP, congratulations! You saw a red "Error 0x18 .... ", my approach is:

(1) Use the Alt + C shortcut or Debug/Unconnect to disconnect CCS from the simulator.

(2) disconnect the DSP power supply

(3) disconnect the USB port of the simulator from the PC.

(4) disconnect the simulator from the DSP

Previously, I often reversed steps (3) (4) and found many problems. Remember, you must first disconnect the DSP and then perform other plugging operations.

2. The V kernel voltage of DSP6000 is always used to measuring whether the power supply and ground are short-circuited with A multimeter before power-on during PCB debugging.

That day, the problem occurred. I used a multimeter (black leather) to test the DSP's V power supply and ground, and the terrible "drop... Drop" sound. It's strange. No way. I am so careful about Circuit Welding. Don't joke. Then I began to check the circuit, continuously add welding, and check every location of 1.2V. I still didn't find any problems. I changed a multimeter (or the black one). "Big Brother", I almost fell down and did not dare to power on again.

At that time, the DSP board was connected with an FPGA. Is it a problem with FPGA? If it breaks down, it will be a short circuit. Then the 1.2V power supply module is also unloaded... short circuit.

Finally, I did not find any problem. I put the Board aside. After a few days, the boss forced me to adjust the board and asked for a multimeter (which is red skin this time, OK. You can take a break.

But science tells me: this is impossible, and short circuits cannot be short-circuited.

Take the original multimeter (black leather) test, "Didi", I was surprised, or short circuit. Red skin test, no short circuit.

Open-minded, power the DSP directly, burn it! Powered on, connected to the simulator, OK, and CCS.

I have to think about this. After the power is down, I tested the resistance between 1.2V and the ground using the Ohm file of the voltmeter, more than 70 ohm (this is the internal resistance of the DSP kernel, a little small ), oh, understand, different multimeter determines the short-circuit threshold value is different, the red multimeter threshold value is lower than the black.

God does not speak, but it does exist silently.