Need help about Cyclone FPGA, too high current!

I've designed a signal acquisiton and processing system based on DSP(tms320c6713) and FPGA(Altera cyclone). When the circuit is powered, the total current is 0.55A though there is no program running neither in FPGA nor in DSP. The other devices, such as A/D, SDRAM and FLASH, do not work, just powered. WHY the current is so high?


Further more, when a simple program is runing in FPGA, the total current is 0.82A~0.83A; when a simple program is runing in DSP, the total current is 0.72A~0.73A.

Can you help me? THANKS!

Hello shishiker,

Do a list of the power consumption for all active parts on your board. This can be done by desoldering the parts or by disconnecting the power pins from the power supply of each part (only possible if your have taken this into account during the layout phase). After that you can make a variance comparison.
There are so many possible failures or perhaps the high power consumption is fine because you are using some parts which needs that amount of power.
It is a good practice for starting up a new board to equip design parts one by one and measure all needed information. E.g. first the power supply, second FPGA, third DSP, …
Did you prevent that two or more IO pins are driving the same line? Did you prevent that there are no floating inputs?


Bye,
cube007

hello, cube007

thanks for your help
It's a pity I've not set switches for every power consumption parts of the circuit, It's my fault. When the circuit is powered, the chips are not hot, just a little warm, except the 5V to 3.3V power convertor is a little hot, and the convertor's output is right.

Later, when I download a FPGA program, there is a new phenomenon.
This time when loading a program, I set all the unused pins as tri-stated inputs. Then when the program runing, the system's current reduced to 0.50A~0.51A. But when there is no program runing, the current come to 0.56A. It's so surprising, as a beginner I can not understand it

can you give me more advice?
Thanks again!

shishiker

Hello shishiker,

Finding a solution for your problem is not easy. Do you have another PCB which isn’t equipped with parts? If yes you can solder first the power supply, FPGA, DSP and so on. After each step measure the power consumption. For a good analysis it is very important to find out which parts of your design is consuming so much power. Did you make a theoretical analysis of the used components? What’s the theoretical power consumption?
From phenomenon in your design it’s easy to say the FPGA is not working correctly. But it’s also possible that there is a other component which isn’t working correctly.
You talked about the 5V and 3.3V power supply. What about the 2.5V power supply for the FPGA core. I think all Cyclones need 2.5V – perhaps that might be wrong.

Bye,
cube007

Hello, cube007

Fortunately, I have other four PCBs, and I'm doing the work you adviced, FPGA and DSP are both difficult to solder for me.

For the theoretical power consumption, just the two SDRAMs in my circuit cost much power (1W for each), but there is no data transmission between DSP and SDRAMs in my work by now. And I've measured the current of the a new PCB only soldered one SDRAM, 12mA~16mA.

The FPGA in my circuit have a 1.5V power supply for the core, and the 3.3V to 1.5V power

convertor is not hot, just a little warm.

Thanks!
shishiker

Hello shishiker,

Sure soldering the DSP and the FPGA isn’t easy. With good equipment you will get good results. For fine pitch parts I use a microscope. With this one you can see all details you need. Perhaps there are some not very well soldered pins on the other PCB which causes shorts between some pins. There are so many reasons …

Good luck,
cube007