question

aeneby avatar image

Tank level sensor update interval is too slow on Venux GX

I have a resistive tank level sensor (similar in design to this one) installed in a header tank, and connected to a Venus GX. It works fine so long as the tank level is relatively static - however, if the tank level is rapidly rising or lowering, the Venus can't seem to keep up.

So for example, if I turn on the pump to fill the header tank, it is overflowing by the time the Venus is even reading 80%. Once I turn the pump off, the Venus reading continues to tick upwards gradually until finally reaching 100%, but this seems to take a few minutes to catch up.

Because I want to use this reading to control the pump, it is obviously an issue. I'm fairly sure it's not the sensor, since connecting a multimeter and actuating the slide manually seems to produce instantaneous change in the resistance reading. Is there a way to get the Venus to update the reading more rapidly?

Venus GX - VGXtank monitor
2 |3000 characters needed characters left characters exceeded

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

4 Answers
aeneby avatar image
aeneby answered ·

Just an update, I have had confirmation off-list that there is a "low-pass" software filter in effect here to filter out the effects of slosh in the tank. Further, there is currently no way to bypass this mechanism.

My personal opinion here is that there are too many assumptions which need to be made (e.g. tank size/shape, sensor position, fill/drain rate, yacht or bus, etc) in order for such a filter to be effective in the majority of cases, and therefore there should at least be an option available to disable it. Perhaps a longer term solution would be to allow modification of the thresholds/parameters in the filter algorithm to enable it to be tuned to individual applications.

1 comment
2 |3000 characters needed characters left characters exceeded

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

If you consider for a moment that there is a feature for pump control, and take into account the very slow software smoothing of levels, you realize that the smoothing as it stands right now makes tank control basically useless or even hazardous.

Since pump control is a configurable trigger from tank level, inadequate response time could allow the pump process to either overflow or run dry the pump.

It would be great to dedicate a bit of time to improving the configurable options in both tank controls and level monitors. Right now it feels like someone checked the checkbox and said "feature complete" without actually determining if the feature was useful.

danhig avatar image
danhig answered ·

Hello,

I just ran into this issue on the final installation phase of a 12 month build. I designed an Ecosystem approach to integrating all of the Cerbo technology (Multi, BMV, MPPT's 4 temp sensors, 3 tank sensors & Cerbo) into a 2020 Winnebago Class B RV to update it from the 1990's class instrumentation as a showcase to what can be achieved with Victron, LifePO4, and a 1500W solar array.

While i am sure this tank monitoring technology works well in large marine vessels its barely usable in an RV and especially with smaller tanks. As an example i start with an empty 13 gallon black tank and put 5.3 gallons of water into it which then updates at a 1% increments 10 minutes. Hows does this help me fill fresh water tanks or all the other use-cases where i need to do something or make a water related decision but i need to ensure no recent activity with the tank/s has occurred. For class B or even C RV's (the largest growing segments in the US after C-19) the Victron implementation is actually no better than the 25% / 4 external stick on pad resolution they have today.

Vicrtron please help me understand when this is not disclosed in your product documentation??? This is a HUGE and expensive 1 way project to dismantle an RV's tank system, drill holes and install the sensors only to find it not any better than what you have today. And there is no going back!

2 |3000 characters needed characters left characters exceeded

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

osaether avatar image
osaether answered ·

Hi @aeneby

I have noticed the same as you but I agree with the above answers that the averaging is generally not a problem.

However, what I don't like is that we cannot select the shape of the tank. In my opinion it should at least be possible to select between the most basic shapes: cylindrical/round and rectangular. I have forked the repo and will look into this.

Maybe a cylindrical shaped tank is not very common on boats, but it is very common in off-grid systems.

Regards,

Ole

2 |3000 characters needed characters left characters exceeded

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

wkirby avatar image
wkirby answered ·

There is some software averaging taking place.
I'd think that it's been designed this way to give a smoother reading to filter out waves inside the tank as the vehicle or vessel moves along. Tanks of a few hundred litres won't fill so quickly, so the gradual update speed is generally not a problem.
I'm sure it will be possible to alter the Venus OS to do what you want. Maybe best to ask on the Venus-dev group: https://groups.google.com/forum/#!forum/victron-dev-venus


2 comments
2 |3000 characters needed characters left characters exceeded

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

This was my conclusion too. Hopefully there is a way to work around this averaging, as it's not appropriate in many cases.

Agreed, it's not appropriate in all cases. And in many where it is, the sensor itself does the averaging in a more configurable way.