CSD23203W: Thermal Hot-spot

Kuramochi Tadahiko

Guru 28715 points

Part Number: CSD23203W

Hi,

My customer is considering to implement the CSD23203W.

They request the thermal hot-spot of this device because their module is very small.

Could you provide this information?

Best Regards,

Kuramochi

over 8 years ago

0 Brett Barr1 over 8 years ago

TI__Mastermind 18200 points

Kuramochi-san,
I was unaware of this type of model. I have forwarded your message to our packaging team to see if they have heard of this.

0 Brett Barr1 over 8 years ago in reply to Brett Barr1

TI__Mastermind 18200 points

For what its worth, the max temperature of this device is 150deg. How much power are you going to be dissipating into it.

0 Kuramochi Tadahiko over 8 years ago in reply to Brett Barr1

Guru 28715 points

Hi Barr-san,

Maximum dissipation is 600mW on customer's condition.

Best Regards,
Kuramochi

0 Brett Barr1 over 8 years ago in reply to Kuramochi Tadahiko

TI__Mastermind 18200 points

Kuramochi-san,

Is that continuous? Even on the min Cu board, the RthJA was only 170degC / W.

Is the FET fully enhanced / turned on during this conduction? If so, they should be fine.

0 Kuramochi Tadahiko over 8 years ago in reply to Brett Barr1

Guru 28715 points

Hi Barr-san,

The customer require the heat-spot where is most hottest point of IC surface.
They use the value to calculate the thermal resistor from hot-spot to each pin.

If you have thermal picture of the CSD23203, could you share it as follows?

Please let me know if it is impossible to provide the point.

Best Regards,
Kuramochi

0 Brett Barr1 over 8 years ago in reply to Kuramochi Tadahiko

TI__Mastermind 18200 points

Kuramochi-san,

I'll let you know what our packaging experts say.

0 Brett Barr1 over 8 years ago in reply to Brett Barr1

TI__Mastermind 18200 points

Kuramochi-san,

This is the response I got from our packaging team: Since the bulk material is Si on CSD23203W, the ∆T within the package is very small. If the pin is close to FET, then the error will be high as well.

0 Kuramochi Tadahiko over 8 years ago in reply to Brett Barr1

Guru 28715 points

Barr-san,

Thank you for cooperation.

> If the pin is close to FET, then the error will be high as well.
Which pin is far from FET?

Best Regards,
Kuramochi

0 Brett Barr1 over 8 years ago in reply to Kuramochi Tadahiko

TI__Mastermind 18200 points

Kuramochi-san,

In this case, because the package is so small, all of the pins are extremely close to the silicon die active area. As such, in order to measure what the thermal impedance from the die to any pin or case will be, you will see very small change in temperature, so very high error in RthJC. As such, you should just use RthJA for your temperature increase (which will be specific to your board but we do give a "worst case scenario" and you can assume the entire FET will be pretty consistent in temperature.

Power management

Power management forum

CSD23203W: Thermal Hot-spot