I am curious about such a high temprature (perhaps more than 50C).I downloaded the same program to both MCF52233DEMO and my PCB.The former is cooler.So I think there must be something wrong with hardware( PCB design).Here's my SCH.Anybody who can find the mistake?
when mcf52235 work normally, its power supply current >329mA
so it will consume 1W.
it is hot if normal. it not hot if it is abnormity.
my 4 layers PCB is cool(65 degree) then 2 layers(75 degree perhaps). because one layer is ground.it help the hot throw out. perhaps carefully layout more ground is a good method.
when open tcp/ip it is hot ,but close it,it is colder(only consume 0.4W).
Thanks,gxd. My chip is not as hot as 65 degree.That is horrible.The temperature ranger is -40 to 85 degree.I can't imagine what will happen running under such a high temperature for 1 year.
I tried the same program with 52233demo and my board. Obviously ,my board is hotter.There's 2 differences . My board has many pins leaving unconnected and it has a different layout (470R pull-up resisters).
The 80-pin QFP MCF52233 on my board runs rather hot when I switch Ethernet on. But I'm assuming this is normal since 1W of power is rather a lot for such a small package. If you plan to run your board at +85 degrees, you should take measures to spread the heat from the MCU. Strictly speaking, it is impossible to run the MCU within the allowed temperature range when ambient is at 85 degrees. There will always be a temperature difference between ambient and MCU. You can only try and reduce it throug the use of heat sinking. Flash data retention time may suffer at higher temperatures. I believe it is specified as 10 years at +85. The high temperatures may affect package reliability too.
I tried a simple program which makes LED blinking with software delay(for(...;...;...)).The ethernet was turned off.However ,the chip was hot too.Perhaps more than 50 degree if I press it for several seconds with my finger.
As you said it is normal. But why the demo board is not hot makes me confused.
Touching the MCU with a finger is not exactly an accurate measurement. Besides, as gxd pointed out, board layout can have an effect on heat sinking. If possible, try measuring supply current to MCU on both boards. That would be more or less an accurate measurement.
I have exatly the same problem. The DEMO board is quite cool but my hardware (which worls fine) has its chip very hot. It is too hot for a finger for more than 10 seconds. The I/Os are not connected (partially populated board) except for the ethernet. When I disable the ethernet the chip runs very cool. I notice no oscillation where it does not belong.
Just a thought... Does the MCF52233 on your PCB still get hot if you configure all of your GPIO pins as inputs? (Or if you download a program that does nothing but halt after reset, and so doesn't configure GPIO pins at all, as they default to inputs?) I'm just wondering if you might have some output pins fighting with the external circuitry...
Thanks a lot! I configged all GPIOs as input.The temperature is a little lower now.But it is still much hotter than the MCF52233demo. I halted the MCU and it cooled down.My PCB is 2-layer board.Isn't there any problem with PCB design?I can't use it with this problem unsolved.