Has anyone potted something like a Raspberry Pi 4

[e100]

The newer models dissipate up to 5 watts so I was wondering if potting would result in overheating problems.

I was planning on using a soft re-entrant silicone gel (https://docs.rs-online.com/532a/0900766b80e0bed4.pdf) with glass beads or sand as a filler to reduce the amount of gel required.

The WiFi module and antenna are on the main board so presumably the performance will be affected to some degree.

[48X24X48X]

While working on some ATEX products recently, resin will help to dissipate the heat better.

[Mr.B]

In my experience, potting compounds like resin and silicone have sufficient capacitance (dielectric) to render the on-board antenna basically useless...
No experience with RPi, but with other WiFi capable devices such as ESP8266 and ESP32.
YMMV.

Edit: to qualify 'dielectric'.

[48X24X48X]

In my experience, potting compounds like resin and silicone have sufficient capacitance (dielectric) to render the on-board antenna basically useless...
No experience with RPi, but with other WiFi capable devices such as ESP8266 and ESP32.
YMMV.

Edit: to qualify 'dielectric'.

I think you can omit the section where the antenna is and it's tracks leading to it.

[e100]

In my experience, potting compounds like resin and silicone have sufficient capacitance (dielectric) to render the on-board antenna basically useless...
No experience with RPi, but with other WiFi capable devices such as ESP8266 and ESP32.
YMMV.

Edit: to qualify 'dielectric'.

I think you can omit the section where the antenna is and it's tracks leading to it.

The metal can over the WiFi chip has some holes so definitely isn't liquid proof.
How do you persuade resin to only go where it's needed? Do you tilt the board and carefully apply resin with a pipette so that it drains away from critical areas?

[e100]

In my experience, potting compounds like resin and silicone have sufficient capacitance (dielectric) to render the on-board antenna basically useless...
No experience with RPi, but with other WiFi capable devices such as ESP8266 and ESP32.
YMMV.

Edit: to qualify 'dielectric'.

Will using a PCB lacquer or any kind of thin conformal coating also have bad effects?

[Mr.B]

Will using a PCB lacquer or any kind of thin conformal coating also have bad effects?

I would expect that it would interfere to some degree also.
The big thing is that you are going to then add another layer of a compound over the antenna traces that will add capacitance.

My solution on an ESP8266 module (sold as ESP-07) was as follows:
I designed the board to have a flat area in a rectangular path around the module and its decoupling capacitors.
I then used a small plastic enclosure with no lid, sold as a potting box.
The box measures about  37mm x 25mm x 12mm (1.5" x 1" x 0.5").
It is placed over the area to be protected from the potting compound and a small bead of silicone is used to glue it down
The module and its antenna are contained in 'free air' inside the box with no chance of silicone or potting compound coming in contact with it.
There are no vias in the board inside the potting box perimeter so that nothing can leak through from the other side of the PCB.
Being a custom designed board allowed me to make the rules...

Of course this is not going to be easy on a RPi board due to other component population, but you may be able to do something.
Keep all potting compounds away from the antenna traces and/or chip antenna.

Good luck.

[cgroen]

Will using a PCB lacquer or any kind of thin conformal coating also have bad effects?

I would expect that it would interfere to some degree also.
The big thing is that you are going to then add another layer of a compound over the antenna traces that will add capacitance.

My solution on an ESP8266 module (sold as ESP-07) was as follows:
I designed the board to have a flat area in a rectangular path around the module and its decoupling capacitors.
I then used a small plastic enclosure with no lid, sold as a potting box.
The box measures about  37mm x 25mm x 12mm (1.5" x 1" x 0.5").
It is placed over the area to be protected from the potting compound and a small bead of silicone is used to glue it down
The module and its antenna are contained in 'free air' inside the box with no chance of silicone or potting compound coming in contact with it.
There are no vias in the board inside the potting box perimeter so that nothing can leak through from the other side of the PCB.
Being a custom designed board allowed me to make the rules...

Of course this is not going to be easy on a RPi board due to other component population, but you may be able to do something.
Keep all potting compounds away from the antenna traces and/or chip antenna.

Good luck.

We do the same trick on 10,000's of devices (GSM/LTE, GPS and 868 MHz antennas). Works fine.

[ejeffrey]

I've heard a lot of people complain about the pi4 in particular having inadequate cooling and strongly benefitting from an extra heatsink and/or fan.

So I would guess any kind of potting isn't going to be great from a thermal perspective.

[salihkanber]

Everything could be potted but a route out must be provided to heat, waves etc. That will also make hardware un-serviceable.
You can use heatsinks or vacuum tubes for transferring heat from the main IC to outside. But that is not the only heat generating component on the circuit