Sealing Issues for Celsius Sensors

At present, we have a PCB board with a Celsius sensor on hand. We want to use this board to package and test it in water. But we don’t know how to seal the sensor reliably. So we want to know how the sensors currently on sale are sealed and waterproof. Especially this kind of PCB board with a more peculiar shape.
If anyone has done related work and has experience with this please help us.

Hi @Franck,

This doesn’t make sense to me. Celsius is the name of our penetrator-based temperature sensor package, and is already waterproof. If you’ve connected one of our Celsius temperature sensors to a custom PCB then you should be able to have your PCB in an enclosure and use the Celsius as a penetrator in an M10 hole, as it’s intended to be used.

If you actually have a TSYS01 sensor on a PCB* (which is what our Celsius uses) and are wanting to know how we create the waterproofing for our Celsius package then I can ask about that internally if you’d like, but I can’t guarantee that’s information we’re able to share, beyond the fact that it’s clearly potted.

*Our PCB schematic and footprint are open source and freely available, and linked in the Technical Details, if that’s relevant.

Yes, I want to seal the PCB with the TSYS01 sensor on it. I’m going to use epoxy to cover and pot this PCB, but often only a thin layer is applied to the PCB to ensure temperature sensitivity. We ended up using other sensor chips, we made our own solution, integrated multiple sensors together, and then put on a sensor every one meter in the Z-axis direction, a string of 10 sensors, and then Ten groups of sensors are used in each of the X and Y axes to form a cubic temperature detection network. Temperature data were measured in matrices of approximately 10 cubic meters, aggregated into a complete data model, and multiple such arrays were used to infer temperature changes across the mariculture area.
Although the way we think of is enough to deal with our work environment, we want to carry out this system in depth and apply it to a deeper environment. So what I’m confused about is whether this approach will work reliably at depths of 500M underwater. Will the water pressure squeeze the epoxy and thus the electronics on the PCB, causing the sensor to be destroyed by pressure rather than by water ingress.
Maybe I should try to use injection molding to wrap a plastic casing around the sensor?

I’m told we use “a thermally conductive epoxy”, and given we rate our Celsius to 975m depth it’s at least possible either

  • to get an epoxy that avoids excessive pressure transfer to the PCB / electronic components, or
  • that the PCB and components are not overly pressure sensitive

I don’t see how injection moulding would specifically help or hinder the results - the main question is around the pressure transfer to the PCB, and whether it is detrimental (which may depend on the stiffness and thickness of the potting material, and likely also its thermal expansion characteristics relative to those of the electronics.