SG9665GC V1 teardown (and some EMI measurements/shielding)

[wozzzzza]

congratulations, you created an oven.

 

[Harsh]

congratulations, you created an oven.

Certainly looks like it.

@Hex Please post temperature readings with the shielding on.

 

[jokiin]

congratulations, you created an oven.

probably more like a faraday cage than an oven

 

[jokiin]

Serial is SGGC15250393. Can you tell me when it was produced?
.

the board is from the first production in early 2015, the assembly was from the second production run, quite a lot of components have been upgraded since that version

 

[wozzzzza]

I got no doubts it will work for what he wants it to do, but a lot of effort and I would like to see the heat affect to it as well, be interesting.

 

[jokiin]

I got no doubts it will work for what he wants it to do, but a lot of effort and I would like to see the heat affect to it as well, be interesting.

it's a huge amount of effort, no denying that, heat is probably ok as he has taken some consideration with that and the copper is possibly helping to spread that over a wider area, none of the vents are blocked which would have been a bigger issue

 

[sp5it]

RFI can be generated by virtually everything

You don't have to explain that. As ham radio operator I smell RFI from a mile

Temperature of processor 69.1 with open case is just 0.9 degree below Operating Ambient Temperature
TOPR -10 ~ 70 manufacturer declare in datasheet.
Imagine how hot it is when case is closed...
IMO some forced cooling would be needed along with cooling fins on top of processor.
Mike

 

[Hex]

congratulations, you created an oven.

Certainly looks like it.
@Hex Please post temperature readings with the shielding on.

Here we go...
Cam was running around 15 minutes, the ambient temperature was 25°C. The boxes show the highest temperature within the images.

And without the cover

As you might see, the copper indeed acts as a heatspreader and with the capton on the copper it can radiate the heat over a much greater surface. Thus the peak temperature at the Novatec processor is reduced from 70 °C to 50°C. So no oven

Tonight I'll do some final tests driving around and checking the DAB reception.

 

[Hex]

Temperature of processor 69.1 with open case is just 0.9 degree below Operating Ambient Temperature
TOPR -10 ~ 70 manufacturer declare in datasheet.

Unfortunately the datasheet only declares ambient temperature and max. storage temperature, nothing you can really work with.
Ambient temperature is obviously the ambient, not the temperature of the processor. Maximum allowed case temperature (or junction temperature together with an thermal resistance when mounted on a decleared board) would be the thing to compare here. Many devices have a max. storage temperature equal to max. junction temperature. When comparing to a similar chip, probably something like an iMX6, you could assume these temperatures:
Input power for the complete cam is around 1,5 W, so assuming 1W for the NT96655 and Max. junction temperatureof 125 °C.
According to the iMX6 Junction to Package Top (what I measure with the termal camera) is 2°C/W, so with measured 70°C the die would operate at 74°C (with a 51 °C margin). Add 30°C for car temperatures (~55°C) the junction would be at 104°C with a 21 °C margin.
With values for Junction to Case of 9°C/W die would operate at 79°C (with a 46 °C margin), and in the car at 109°C with 16°C margin.

Of course, this are all assumptions and maybe totally wrong...

iMX6 Datasheet: http://www.nxp.com/docs/pcn_attachments/16116_IMX6SLCEC.pdf
A nice paper from Macronix about this topic: http://www.macronix.com/Lists/Appli...Plastic Package Device Thermal Resistance.pdf

 

[sp5it]

So what is ambient temperature inside camera while processor is 69 and diode is 78 on idle when opened?
Can you put NTC inside, close case and measure that?
Mike

 

[Hex]

So what is ambient temperature inside camera while processor is 69 and diode is 78 on idle when opened?

Already documented here:

The measurements are taken at room temperature (~23°C) and when the camera was running a few minutes.

The ambient temperature inside camera was obviously also room temperature.

Can you put NTC inside, close case and measure that?

No NTCs, only type-K thermocouples, but it shouldn't matter
Because I wanted to know which temperatures the cam must withstand when parked in the sun I put the thermocouple trough the venting holes above the SD-Card inside, the other measurement was the air temperature a few cm away from the cam. Environment: Outside temp in the afternoon: 33°C, windows of the parked car are opened a few centimeters, AC was on and set to 22 °C.

Drive in the morning
Start
Temperature near the camera: 25 °C
Temperature inside the camera: 25 °C

Stop (after 15 min drive)
Temperature near the camera: 23 °C
Temperature inside the camera: 31 °C

Drive in the afternoon
Start
Temperature near the camera: 54°C
Temperature inside the camera: 54°C

Stop (after 15 min drive)
Temperature near the camera: 23°C
Temperature inside the camera: 35°C

So it looks like I forgot one important point regarding cooling the camera: Ventilation and AC from the car.
It seems like the cam is getting actively cooled inside the car

 

[jokiin]

Drive in the afternoon
Start
Temperature near the camera: 54°C
Temperature inside the camera: 54°C

Stop (after 15 min drive)
Temperature near the camera: 23°C
Temperature inside the camera: 35°C

So it looks like I forgot one important point regarding cooling the camera: Ventilation and AC from the car.
It seems like the cam is getting actively cooled inside the car

put your hand on the glass outside before you drive off and then check it again after a driving a while and you'll feel a difference, even without AC just driving will also cool the window which has a carry on effect