We have a new build narrowboat and the fresh water tank level has been problematic since delivery. It swings around by 100-150 litres based on time of day with a capacity of 800 litres. We coped with it, but now the boat is laid up for winter can see it is just moving daily - and knowing correlation <> causation - it appears to be falling as system voltage rises when solar cuts in then recovers when the solar drops off. Most times overnight the signal is stable when the batteries sit at a single value for 10-12 hours. The shot below shows the past 7 days (the behaviour has repeated for 6 weeks, the chart is just too hard to read):
I don’t know the sensor fitted but think it is a 4-20mA pressure sensor from the build documents. It is connected to a GX Tank 140 like our fuel and black water. Fuel is very stable and black water shows similar but not as large amplitude swings. My limited searching on the topic seems to point to a ground reference issue but would really appreciate any insights or similar experience (+ Fixes?) the community may have?
Hi Darren – can you also plot temperature as well as voltage? I assume you’re running with solar/wind priority enabled so the voltage rises up during the day (solar power) and drops back at night (dark), so maybe daily temperature changes rather than voltage is causing the variation? (though I don’t really see how for a water tank sensor at the bottom of the boat…)
It would be strange for the main battery bank voltage to affect the reading, the Tank 140 is powered via USB and generates the 24V needed to power the sensor from this and talks to the Cerbo via the same USB interface, and this must be working OK.
Is it a Float sensor (rises with height of the water in the tank) or a Pressure sensor (kPA or PSI)? Suggest you have a look for the make/model of sensor and data, as well as have a look at where it is installed.
Ian is on to something here with the temperature. My recent experience, new build narrowboat in 2021 with MCS Boat Products level gauges on water, diesel and black water linked to a GX Tank 140, although these are 0-10V voltage signal not 4-20mA current and work on pressure. These have worked fine for over 5 years, rock solid on all 3, never a problem. This summer in July the water tank started showing cycling behaviour and the only thing I could link it to was temperature. I plotted out the cabin temperature as I have a sensor on the GX for this against the water level for a period the boat was out of use. There is no way that the water level was actually fluctuating this much with temperature when it had never showed it before. In the end, the only solution was a new sensor and now the level is once again perfectly well behaved with temperature fluctuations. I can only guess that something went awry with the circuitry in the pressure sensor, because if it was with the GX Tank 140 it would have carried on with the new sensor. The other levels were not affected. Mine was much more sinusoidal than yours, which does follow the voltage more. Sorry I can not be of more help.
I expect it’s the same as mine, a 0.1bar 4mA-20mA pressure sensor (5mA/ft of water) fitted onto a spigot at the bottom of the water tank (so output flow doesn’t affect reading), and connected to a Tank 140.
Mine shows a much smaller variation, see attached plot – the occasional bigger spikes are when there’s somebody on the boat, any tilting of the boat affects the reading because the tank is across the width of the boat but the sensor is at one side. I believe Darren’s tank is shallower and wider than mine, but that doesn’t explain the huge difference in variation, especially since there’s nobody on board his boat.
Another thought – I believe those sensors report the difference in pressure between the front (fluid) and the back (atmospheric pressure) of the sensor, there’s usually a small vent for this. If this vent is blocked then the reading will also be sensitive to both atmospheric pressure (only seen on the water side) and probably temperature.
Thanks Ian,
I don’t think there is a temperature data stream that shows daily changes - all I can find is the battery one and that is much more moderated - although probably a good proxy for what the water tank sees. I originally thought temperature but then realised it is cycling with the solar so e.g. when the BMS stops charging around 2pm the tank level begins to “recover”. If it was tracking temperature I think it would do that “recover” phase later in the day. I suspect we do have the same sensor and I know our tank is wider and shallower than yours.
Managed to find an external source for temperature - net result is a regression suggests a much weaker explanatory variable than the voltage. Will try to see if I can get it looked at as it’s moving upto 20% a day at the moment which is not ok. Appreciate the suggestions - thanks
This is all very strange – given how the 4mA-20mA sensor connects to the Tank 140 (including power via USB, and immunity to GND voltage errors) it’s difficult to see how the main 48V battery voltage can affect the tank level reading, since it never gets anywhere near any of the circuits…
Pressure sensors come in 2 types, gauge (relative to atmospheric) and absolute (relative to vacuum). For boat tank measurement like yours you want a 0.1bar gauge sensor, this will measure water depths from 0 (4mA) to 3.2ft (20mA), sensitivity is 5mA/ft – from looking at the sensor setup in VRM, I know this is what I’ve got.
They have a small vent hole to the rear of the diaphragm, if this is blocked (e.g. dirt/oil) then they’ll also be sensitive to atmospheric pressure (on front of diaphragm) and temperature (closed air on rear of diaphragm). This could be your problem.
A second but unlikely possibility is that an absolute sensor is fitted but this is very unlikely, it would have to be at least a 2bar sensor and would only have 0.5mA/ft sensitivity as well as being sensitive to atmospheric pressure.
