Post History
I guess it would work. I have used electrodes on a PCB to detect whether water level got high enough to turn on a sump pump. It worked, although it hasn't been installed very long. In that case ...
Answer
#2: Post edited
by
Olin Lathrop
·
2024-02-20T17:44:21Z (10 months ago)
- I guess it would work.
- I have used electrodes on a PCB to detect whether water level got high enough to turn on a sump pump. It worked, although it hasn't been installed very long. In that case the PCB extends upwards to where it is dry. That's where wires are soldered that go to the detection circuitry. The exposed parts of the electrodes are gold plated. I don't think there is anything else you can commonly get from a PCB process that would work.
- I once worked on a product that had to detect whether it got immersed in a glass of drinking water. In that case, they used platinum-plated electrodes that were like thick and stiff wires coming out of the unit thru rubber grommets, so no PCB material was abused.
- If I were doing this in a real product, I'd start with what you suggested, then try to do accelerated aging tests. I don't know what happens to soldermask after years of contact with water. It <i>seems</i> like it should be OK, but I'd test it before committing to volume production.
- One thing I noticed with my one-off sump pump switch is that a film of crud gets deposited over anything immersed in the water for long. It doesn't hurt the electrodes, but I don't know if long term it might make it difficult to tell wet from dry due to the leakage. My particular water has a lot of iron in it.
One of my current investigations is sensing water level (not just present or not present) with a PCB, ultimately intended for a real product. I am uncomfortable with a direct electrical connection, so am using capacitive sense. Everything is covered by the soldermask. The one-off prototype is working very well, but I have no long-term reliability info.
- I guess it would work.
- I have used electrodes on a PCB to detect whether water level got high enough to turn on a sump pump. It worked, although it hasn't been installed very long. In that case the PCB extends upwards to where it is dry. That's where wires are soldered that go to the detection circuitry. The exposed parts of the electrodes are gold plated. I don't think there is anything else you can commonly get from a PCB process that would work.
- I once worked on a product that had to detect whether it got immersed in a glass of drinking water. In that case, they used platinum-plated electrodes that were like thick and stiff wires coming out of the unit thru rubber grommets, so no PCB material was abused.
- If I were doing this in a real product, I'd start with what you suggested, then try to do accelerated aging tests. I don't know what happens to soldermask after years of contact with water. It <i>seems</i> like it should be OK, but I'd test it before committing to volume production.
- One thing I noticed with my one-off sump pump switch is that a film of crud gets deposited over anything immersed in the water for long. It doesn't hurt the electrodes, but I don't know if long term it might make it difficult to tell wet from dry due to the leakage. My particular water has a lot of iron in it.
- One of my current investigations is sensing water level (not just present or not present) with a PCB, ultimately intended for a real product. I am uncomfortable with a direct electrical connection, so am using capacitive sense. Everything is covered by the soldermask. The one-off prototype is working very well, but I have no long-term reliability info.
- <blockquote>You mentioned platinum-plated electrodes which dipped into drinking water. I noticed that high-end conductivity sensors use platinum electrodes (low-end sensors use stainless steel). I wonder why platinum? Is there something wrong with gold?</blockquote>
- I don't know. I suspect two issues, process and durability. There is a well established process for applying gold coatings to PC board conductors. You can get gold-plated wire, but it's not common. We use gold-plated nichrome wire in some of our products, and it's a custom special order and not easy to arrange.
- Gold seems better for contacts. I've tested gold/gold and gold/nichrome contacts. Even though nichrome is reasonably inert and it was clean, the electrical contact was less reliable at low contact force.
- On the other had, I expect platinum to be more mechanically durable.
- However, this is all speculation and the original answer stands: I don't know.
#1: Initial revision
by
Olin Lathrop
·
2024-02-19T21:08:04Z (10 months ago)
I guess it would work. I have used electrodes on a PCB to detect whether water level got high enough to turn on a sump pump. It worked, although it hasn't been installed very long. In that case the PCB extends upwards to where it is dry. That's where wires are soldered that go to the detection circuitry. The exposed parts of the electrodes are gold plated. I don't think there is anything else you can commonly get from a PCB process that would work. I once worked on a product that had to detect whether it got immersed in a glass of drinking water. In that case, they used platinum-plated electrodes that were like thick and stiff wires coming out of the unit thru rubber grommets, so no PCB material was abused. If I were doing this in a real product, I'd start with what you suggested, then try to do accelerated aging tests. I don't know what happens to soldermask after years of contact with water. It <i>seems</i> like it should be OK, but I'd test it before committing to volume production. One thing I noticed with my one-off sump pump switch is that a film of crud gets deposited over anything immersed in the water for long. It doesn't hurt the electrodes, but I don't know if long term it might make it difficult to tell wet from dry due to the leakage. My particular water has a lot of iron in it. One of my current investigations is sensing water level (not just present or not present) with a PCB, ultimately intended for a real product. I am uncomfortable with a direct electrical connection, so am using capacitive sense. Everything is covered by the soldermask. The one-off prototype is working very well, but I have no long-term reliability info.