Quoted from "facboy@gmail.com"

virtual flow sensor is not very accurate

it is an indictor if you have flow or not.

leakshield requires an accurate measurement so that the pressure can be adjusted accordingly.
But it is also no problem to do this without a flow sensor.

Quoted from "facboy@gmail.com"

is that all leakshield would use it for? i.e. a binary flow yes/no?
i guess i'm not very keen to block the flow in the loop while the pump is running (even though presumably that is how the pump has been designed), it just seems...wrong.

I have a test loop using 10mm/16mm soft tubing, I tried calibrating my D5 Next Flow measurement by kinking a tube but this resulted in a flow rate that was way lower than what my High Flow Next reports. I read somewhere that the calibration works better if you cut off the flow with a ball valve. I installed a ball valve in the loop and did the calibration again. The new flow rate reported was closer to what the High Flow Next reports but still significantly lower. I think this is due to the accuracy limitations.

As for intentionally stopping the flow, I don't like doing it either. I think blocking the flow is different than having no coolant in the pump. It seemed to run OK with the ball valve closed but it still made me uncomfortable. I tried searching for whether its OK to run a D5 pump with the flow blocked but could not find anything. I think if the flow is blocked, the pressure will increase and I guess its OK to do this for 30 seconds or Aquacomputer would not say to do this as part of the calibration process.

My high flow NEXT an D5 NEXT deviate by 22 - 30 %, 101 l/h from the High flow NEXT and 122 - 130 l/h from the D5 NEXT if i set the correction factor within the flow menu of the D5 to -25%, otherwise it shows 163 - 171 l/h at 0% correction factor.
It isn´t accurate at all.