Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

The theraml analysis in IC design

Status
Not open for further replies.

chang830

Full Member level 5
Joined
Feb 11, 2006
Messages
267
Helped
14
Reputation
28
Reaction score
3
Trophy points
1,298
Activity points
3,424
Hi,
The circuit I am designing need to have hgh drive capability. In layout, I put the driver transistors compact to inprove the matching charateristics. But I found the thermal intensity is high to 200W/cm^2 for the area of compact driver transistors.

I have the concern the heat will brings the reliabilty for my IC.

Any guideline on the restriction for the heat on chip?

Thanks
 

Try and put a temperature detection loop for the circuit so that I can operate in the lower current mode when the temperature exceeds a certain value.

Layout wise, you can actually try and spread the transistor to different corners of the chip. But this will cause a routing overhead and also a substantial IR drop if thick metal is not used.

Externally please put the provision for a heat sink for the place where the drivers are.

Added after 7 minutes:

By the way, are you sure about the numbers because 200W/cm^2 is hot bulb numbers.......
 

    chang830

    Points: 2
    Helpful Answer Positive Rating
Vamsi Mocherla said:
By the way, are you sure about the numbers because 200W/cm^2 is hot bulb numbers.......

Hi Vamsi,
Thanks for the reply.It is very helpful to me.

As for the 200W/cm^2 power density, that's waht I am being confused in present.

My 1st question is what is the uplimit for power density for convetional silicon chip?

Second, how to compute the power density in design phase? I used the current throught the transistor divided by the area the transistor occupying, which leads high up to surprisingly 200W/cm^2. Any wrong with it?

Thanks in advance for the help!
 

Dear Chang,

Typical Modern microprocessors take about 45 -65 W/cm^2 in today's technologies. Refer to Intel's Power group which has all the facts on their websites. Another good reference is **broken link removed**
That was the reason why I said that your circuit is a hot bulb.

May be you must have computed by the equation power density = V*I/area of the transistor., where V is the power supply voltage and I the rms current. You should also try to factor in the duty cycle of switching in case of a switch or the 2*Vdsat of the the device, if it is operating in saturation.

I am sure that your pass transistors or power transistors in a voltage regulator will have higher power densities. Hence they are laid out at the periphery of the chip to alleviate thermal issues and IR drops.

I hope this helps
 

    chang830

    Points: 2
    Helpful Answer Positive Rating
I think larger pass element will be fine.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top