We ordered 2 sets of the SPEblox and currently testing it with the GIGeblox for our application.
We are seeing the SPEblox getting pretty hot on 2 chips. Can you provide the thermal power for these chips and is there any recommended heatsinks we can use to keep them cool.
Our application demands integrating these boards into a fanless enclosure that is IP54 rated for industrial applications. We would appreciate any guidance on thermal and mechanical mounting
SPEBlox does tend to get quite hot, which does somewhat limit operating temperature. The first way to solve this is to reduce the input voltage. Running the board at 10V will reduce the heat generated by the onboard DCDC regulator.
However, most of the heat comes from the two PHY chips on the board, U2 and U7.
Thermal specifications for both chips are below.
U2
Maximum operating temperature: +105°C
Chip size: 7x7mm
Max power dissipation: 537.2mW
Junction to case thermal resistance: 18.7 °C/W
U7
Maximum operating temperature: +125°C
Chip size: 6x6mm
Max power dissipation: 679mW
Junction to case thermal resistance: 22.2 °C/W
In this case, some passive cooling to help the chips dissipate the heat into the air is going to be enough to reduce the cooling. A good choice would be the V5618C, which is a 23mm x 6.35mm heatsink that can be placed across both chips.
You would need either some type of thermal tape (like this) or a thermal epoxy (like this) if your application is going to be subject to vibration.
Hello,
I’m noticing considerably high heat coming from Y1 (52°C to 54°C). U2 is only at about 41°C, while U7 is only at 24°C. I’m seeing this on two boards connected together. Is there something wrong with my boards?
NOTE: I’m the same person seeing the 323/338 kHz noise + fundamentals who just posted in the, “SPEBlox 1000BASE-T1 master vs slave”, forum.
High heat from Y1 is unexpected. That’s a simple crystal oscillator and dissipates on the order of milliwatts. Are you sure it’s coming from Y2 and not U2? How are you checking the temperature?
Are the boards actually working/passing data?
Hi Josh,
I’m aiming a IR gun directly at the chip, and at the oscillator.
Something I just noticed…Y2 is hot on the slave board, and Y1 is hot on the master, and I’m noticing the complimentary chip (U2 for Y1 and U7 for Y2) are also hot.
Using an input V of 12 VDC by the way.
We’re now approaching 60°C on the slave unit.
I suspect your IR gun is not fully accurate. Y2 and Y1 dissipate very little power. It’s probably the main chips that are generating the heat, but Y1/Y2 have metal cases so perhaps show up more on the IR gun because they are reflective.
Again, none of this is particularly out of the ordinary. The boards are super small, and they do tend to heat up a fair bit.
HI Josh,
You’re probably right on the Y1/Y2 theory. I’m using a cheap Milwaukee IR gun, not a lab grade device.
Thank you,
Joe
