Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
14 years ago

Can Cyclone III Be Damaged By Heat?

I couldn't find any good information on this through Google or searching these forums, so hopefully this hasn't been answered before.

Is it possible for a design to damage a Cyclone III device (C120, dev kit) with heat? I have a design that uses 90K LEs, runs at 50MHz, and most of the logic is switching every cycle. When I run it, the FPGA gets hot enough to burn my finger. I don't have a temperature sensor of any kind, so I don't know what the actual temperature is. It draws about 2.5amps according to the reading on the dev board.

So, can the chip be damaged from this? If so, what can I do to prevent damage? Will a fan suffice, or does it need an actual heat-sink?

Thank you!

2 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    14 years ago

    Thank you for the response thepancake! I indeed had not looked at that. I tried it out for a bit, but had trouble figuring out how to use it. I saw that it referenced PowerPlay in Quartus, something which I hadn't used before. So I tried using that. A couple of questions:

    Will PowerPlay give me an appropriate estimate of the chip's temperature? I saw that it listed an estimate of junction temperature, so that, I assume, is the value I should be looking at.

    PowerPlay was setup, it seems by default, to expect a heatsink on the chip. I changed that setting to no heatsink and no fan, since there are none on the C120 Dev Kit. Was that the correct thing to do?

    PowerPlay's estimate of power usage was way off, compared to what the on-board power meter tells me. PowerPlay reported that the toggle rate was probably wrong, and a simulation should be run. How do I resolve that issue?

    Would it be a better idea to use this equation (from WikiPedia): 

     TJ = TA + ( R θJA × PD )

    and use the power reading from the on-board power meter for Pd? Would that give a better estimation of the real-world temperature, vs. PowerPlay's estimations?

    Sorry if these questions seem simple. I will be browsing Altera's documentation to try and answer them myself, but I figured it was worth asking in case I can't find the answer.

    Thank you for your help!