Slightly off Topic

Apologies for the off topic post, but I know there some very knowlegable people in this forum that might know the answer.

Do any of you know much about CMOS sensor ICs. I am having to do some thermal calculation on a system that has one of this implimented. Its a high end CMOS sensor, that draws 4w of power. The only problem if, when I model this at 4w the temperature gets crazily hot (in free air), but in reality it only gets 40c above ambient at the 4w.

Im therefore wonder if all 4w is actually converted in to heat, or maybe CMOS / CCD sensors are different in the way they use the power and in fact a lot less is converted in to heat?

I would suggest that the sensor doesn't actually draw enough current at the supply voltage to generate anything like 4W of heat. Where's the 4W figure coming from? If it's from the "Absolute Maximum" part of the data sheet then it will most probably be way higher than the actual sensor power use under normal conditions.

Best bet is to measure the actual voltage and power supplied to the sensor and calculate the true power. My guess is that this will be a lot lower than the 4W rating.

As an example, I've just finished potting up a CMOS camera chip and board. The rating for that suggests that it uses around 300mW. In reality it uses around 100mW, if that.

Hi Jeremy,

Thakns for the reply.

The 4w is what the manufacturer advised me it draws (Its a high end CMOS that measures at a very fast frame rate). Unfortunately I am not able to measure the current draw.

RustyH said:

The 4w is what the manufacturer advised me it draws (Its a high end CMOS that measures at a very fast frame rate).

Click to expand...

All the energy must go somewhere. If there is no energy going anywhere, it is not drawing power.

But shouldn't the thermal calculation of the system use the maximum anyway? If 4W is specified as the maximum, then doesn't that mean that it could use 4W under some conditions? Note that loads attached to the sensor could also affect it.

RustyH said:

Unfortunately I am not able to measure the current draw.

Click to expand...

Why not?

nick12ab said:

All the energy must go somewhere. If there is no energy going anywhere, it is not drawing power.

But shouldn't the thermal calculation of the system use the maximum anyway? If 4W is specified as the maximum, then doesn't that mean that it could use 4W under some conditions? Note that loads attached to the sensor could also affect it.

Click to expand...

I agree, but the issue I have is if I set the thermal simulation to 4w, the CMOS gets to around 450c in free air, which doesnt reflect to what the manufacturer says that it should be only +40C hotter at 4w.

If to achieve +40C, I would need to drop the power being dissipated in heat to quite alot lower (like 1w or so)

nick12ab said:

Why not?

Click to expand...

Its a £20k camera system that I would not be authorised to strip and poke at

RustyH said:

the issue I have is if I set the thermal simulation to 4w, the CMOS gets to around 450c in free air, which doesnt reflect to what the manufacturer says that it should be only +40C hotter at 4w

Click to expand...

Is this not an issue with the simulation or the simulation software ?

4W = 800mA @ 5V. You could see what effect putting a 6.25R resistor across 5V has which should also draw 800mA and be 4W.

4W is about a third of a 60-watt-equivalent Compact Florescent Light bulb (120V in the U.S.). Such a bulb will get warm, maybe even hot, but not crazy hot.

If there's a device of any significant mass which will develop 450C above ambient temperature on 4W, the electrical heating industry will be revolutionized.

On the other hand, I wouldn't want to run 4W through a 1/8th watt resistor.

lbenson said:

4W is about a third of a 60-watt-equivalent Compact Florescent Light bulb (120V in the U.S.). Such a bulb will get warm, maybe even hot, but not crazy hot.

If there's a device of any significant mass which will develop 450C above ambient temperature on 4W, the electrical heating industry will be revolutionized.

On the other hand, I wouldn't want to run 4W through a 1/8th watt resistor.

Click to expand...

We are talking junction temperature here, and in a bulb the junciton I guess would be the filament.....which gets extreamly hot to create the light. But I do see where your coming from, 4w creating that high a temp doesnt seem right does it......Im taking a good look at the model setup

RustyH said:

I agree, but the issue I have is if I set the thermal simulation to 4w, the CMOS gets to around 450c in free air, which doesnt reflect to what the manufacturer says that it should be only +40C hotter at 4w.

Click to expand...

Does the simulation model include the correct thermal conductivity with the mounting, other parts and environment, etc.?