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 temperature effects on clock speed

Status
Not open for further replies.

no_mad

Full Member level 5
Joined
Dec 10, 2004
Messages
271
Helped
30
Reputation
60
Reaction score
11
Trophy points
1,298
Location
Naboo
Activity points
2,489
Hi all,

I have been asked in interview regarding the temperature effect on
clock speed (max clock frequency) of a circuit.

1) If temperature increases, how does it effect the clock speed? Will
it increases or decreases clock speed?

2) If voltage source increases, will it increases or decreases clock
speed?

Please share your thought on this topic.

Thanks in advance.
 

Re: Temperature Effect

If you mean to say clock speed is max clock frequency.

Then it is going to reduce with increasing in temperature, because cell delay will increase with increasing in temperature.

And max clock frequency will increase by increasing in power sources, because IR drop will be less so cell delay will be less. But increase in frequency will not be by significant amount.
 

Re: Temperature Effect

To find out the effect of temperature on clock speed we have to look into semiconductor physics. Semiconductors ( p-type and n-type) may exhibit either +ve or -ve temperature coefficient. When temperature increases, the mobility in MOS transistors may decrease. This decrease in mobility increases the Vt of the transistor. This makes the transistor slower. Thus, an increase in temperature will decrease the clock speed of a digital ckt.

Increasing the voltage source increases the potential difference at the gate and also if the source and drain voltage difference is high, the number of carriers injected may also increase. This increase in carrier injection will tend to increase the switching speed of the transistor thereby increasing the overall clock speed of the digital ckt.

Correct me if I'm wrong here!!
 
Temperature Effect

Another factor which also contribute to cell delay to increase is resistivity.

When temperature increases, this increases resistivity.
Am I correct?
 

Temperature Effect

That's true for semiconductors but not for metals. That mean that increasing temperature will reduce delays in medal interconnect.
I have a question:
When Temperature rise Transistor becames slower however interconnects becames quick. Which phenomenon will dominate ? The system will be faster or slower when Temp. rises ?
 

Re: Temperature Effect

yes there is an effect on clock frequency due to changing in temperature. you could read it in a datasheet it will state ppm (part per million). i attach a graph for your review
 

Temperature Effect

with the increase in Temp, the operating voltage will be lesser ( chk ur .lib file) and hence ur design will be characterized for slower operating conditions !

this is bcos, in the "slow.lib" the
1) operating voltage is "low"
2) operating temp is "high" +80c

hence all the max delays of the cells, nets wil be taken while doin synth and STA. therfore resulting in a reduced operatin freq !!
 

Re: Temperature Effect

OK, now I have another doubt.

From my reading, when

temperature increases
---> Vth increases (according to vlsi_whiz)
When temperature increases, the mobility in MOS transistors may decrease. This decrease in mobility increases the Vt of the transistor.

If I'm not mistaken, when temperature increases, Vth of a transistor will decrease.
Thus, lead to my next doubt.

---> transistor become slower (according to master_picengineer)
When Temperature rise Transistor becames slower
Would you or anyone please explain how or which parameter make a transistor become slower when temperature increases.

What I know.
If I'm using worst_case lib during synthesis, my max clk freq will decrease.
worst_case or slowest lib --> voltage low, temp high, slowest process

Thus, can I make a conclusion that if temp high, max clk freq decreases.

I'm confuse..:?:
Pls help me
 

Re: Temperature Effect

Vth is the function of the Fermi potential. The Fermi potential is proportional to the Temperature. We can said that, when temperature increased, the Vth also increase. If other parameters constant, the current will decrease if Vth increased.

For ring oscillator, oscillation frequency is proportional to the current.
 
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top