Loop Flow Rates

[james1701]

I have noticed since setting up my loop loop using an MPS 400 sensor, my pumps are set at 2.5. Over the last two weeks, every day my flow drops a little. When I initialy set it up, I was running around 355l/h. Now with no changes, I am getting about 310l/h. Everything was flushed prior to install. Temps are still good and holding steady.

Is it normal for flow to decease a little over time in a new loop as components settle in, or could it be the sensor?

 

[Tweakin]

Excluding gunk build-up within the system from dyes and such, I would think the flow rate would increase as all the micro-bubbles of air are removed from the system.

Outstanding question.

 

[james1701]

No dyes, just distilled and kill coil and one drop of dead water.

 

[Tweakin]

Maybe my thinking is flawed. I have never used a flow meter, I just monitor the rpm of the pump, and I have noticed that the rpm will drop during the first week or so of usage, and rpm is directly tied to flow rate.

Does your sensor have one of those paddle wheels inside to determine the flow? I have seen posts that the paddle runs on needle bearings that can deteriorate causing the wheel to slow down. One post even stated that the wheel froze causing the entire pump and video cards to burn up. Not sure If I buy that or not, but I see some logic to the statement.

 

[james1701]

No wheel, its digital.

 

[Tweakin]

I just looked it up...nice! I will still fall back on my original assumptions that flow should increase until the micro-bubbles are all gone from the system.

However, I'm going to do some additional research and see what else I can find as I'm curious now...

 

[john3850]

I have a few old EHEIM 1250 317 GPH pumps that do not have the pressure to work with any decent wb.
Replaced the EHEIM 1250 317 GPH with any MCP655 pump and there was no more pressure problems.
Try to measure your pump volts or draw if possible there could also be some blockage or junk that was loose on the parts.
Last problem I had was lint build up in a new wb.

 

[Tweakin]

I have a few old EHEIM 1250 317 GPH pumps that do not have the pressure to work with any decent wb.
Replaced the EHEIM 1250 317 GPH with any MCP655 pump and there was no more pressure problems.
With a closed impeller such as the MCP655 uses puts out more pressure at the same rpm then open style impellers do.
Flow meter = volume not pressure.

Not sure what you are referring to with the pressure issue. Regardless, is what the OP is seeing considered normal operation?

 

[james1701]

I am running dual D5's. Although I wish I had put a pressure sensor inline. Instead of just reading full or empty on the res, I could look as see if it would correlate to increased pressure equal a lower flow rate. That would indicate blockages in the system.

I might want to pull my res out and crack the top cap. When I lowered my water level to flip my flow meter, I had some air that was trapped because of too much water in the res causing a nasty pump hum. When that air escaped the pump went to silent normal operation. I wonder if that little extra pressure in the res is enough to restrict flow rates.

 

[Rubycon]

A few things to clear up confusion here.
Your transducer may have a digital interface /output but the actual sensor is measuring an analog value. Just as a microphone measures SPL, this information can be converted to a digital value, etc.

Most transducers in the hobby are basically proof of flow devices, accuracy or repeatability isn't guaranteed for sure! Just as you cannot measure electrical current alone to deduce wattage, a flow value from a sensor cannot be used to (accurately) determine fluidic flow in a circuit without a comparator which has a value of pressure simultaneously.

Why does the value diminish? Could be a variety of things. The sensor itself could need a break in period. Some sensors employing piezo-resistive bridge transducers are also affected by temperature. If the transducer itself isn't temperature stabilized or at least compensated its value will drift sometimes substantially with changes in ambient temperature.

If the sensor employs the common method of using a spinning disc or wheel with a tiny neodymium magnet that triggers a reed or hall effect transistor, its bearings can change slightly over time. Friction will always increase resulting in a lower rotational rate. A pressure transducer at the pump's volute can also be used as another metric to see if lift load has actually changed. Shift in coolant specific gravity will always elevate this value. Dissolved gases make minute shifts as well.

The electronics used to count pulses or measure differentials across a bridge will also have tolerances to be considered. If the moving element is visible on the (proof of flow) device a stroboscope can accurately determine its rotational speed. Change in speed without change in pressure/viscosity/sg/temperature could suggest increased friction developing in the bearings or fouling, etc.

I would not lose sleep over this unless the value has a chronic drop over a period of time with no end in sight.

 

[james1701]

Ruby, thanks a bunch.

After more research I found this particular sensor uses a differential pressure measurement. The way it gets its information take data from a pressure sensor that gets converted by a segmented calibration algorithm into a flow rate.

The out of the box calibration has been reported as fair, but not entirely accurate. The sensor is very good, and has the capacity for excellent reporting abilities, with a proper calibration. I have yet to learn how to do that for myself, yet.

After sifting through the manual more, I found a line that said the unit needed regular zero flow calibrations for accurate reporting. The Aquasuite software does have the ability to automatically zero the flow when everything is shut down, but power is still running through the USB port to the sensor.

So instead of taking the sides down to unplug my pumps, I made computer to go to sleep. I punched a few buttons on the Aquaero to verify it was still working even though the computer was asleep. After I woke it up, my flow rates were almost back to normal, 338l/h.

I'll keep an eye on it and see how fast it drifts back out of shape.

 

[Tweakin]

A few things to clear up confusion here.

As always, you bring clarity to an otherwise chaotic situation! Nice to see you around the forums again.

 

[Rubycon]

It's a differential sensor which is a good start. Definitely needs a calibration as out of the box isn't going to be accurate however all things said, it should be somewhat relative.

Somewhat. Usually on those sensors you have several points to cover with a span to get a wider range accuracy of +/- 2% (for example).

Also, the electronics on the head (Aquaero?) are likely to drift since there is no oven but there may be anticipation built in. Just make a note of the readings, they should stabilize and then go from there.

 

[SimsReaper]

Everything Rubycon said is correct. But I would also like to add a few other factors here as well. I'm an Instrument Technician by day, and as such have a lot of experience and knowledge regarding flow meters, pressure sensors, etc...

The MPS400 you have uses differential pressure measurements to give a flow rate. The only way this can work is that the MPS400 has to have a fairly significant flow restriction itself in order to obtain the two separate pressure readings. I will provided a picture to help explain.

Your readings from your flow meter will be heavily influenced by any turbulence in your loop, especially the presence of microbubbles in the flow, as the will have a noticeable effect on the pressure drop after the restriction device (orifice plate). Another component that introduces turbulence into your meter is the arrangement of the tubing/ piping. For a meter like this to be accurate, you should have it installed in a place where there is a straight line for at least a few inches before and after the meter. This assists in achieving "laminar" flow, and reduces eddies in the flow which can give erroneous readings.

As for your original question, I wouldn't worry about a small decrease in flow rates, as this could very well be attributable to the settling out of microbubbles and the wearing in of the pump. As Tweakin and Rubycon said, then you can worry about it.

As for calibration, to achieve this normally you would need a separate flow meter that is already calibrated and installed in the loop, and could then find a calibration factor to offset your readings to the correct ones. I do not know the MPS400 software suite, it may have a different process, but either way you would need accurately calibrated gauges/ meters for this to be viable.

The only real use for a flow meter in a loop like this is as a proof of flow, loose troubleshooting tool. You should never rely on the meter for accuracy in measurement, but it should be good for +/- 10%, depending on the installation, and the quality of the sensors used. What it will tell you is that your system is flowing well, at what pressure it is flowing at, (at the point of installation at least), and it should allow you to plot over time any loss of flow.

Hope this helps a bit with the understanding of the meter itself if nothing else. You should understand though, the use of a differential pressure to determine flow does so with a significant restriction to flow.

Cheers, and good luck with your loop:thumbsup:

 

[Rubycon]

Good points on the high pressure drop.

Also, microbubbles is something that should never happen in a PC cooling circuit! When filling for the first time, the pump should only be turned on briefly and as soon as you hear its sound change, disconnect power immediately! I cannot stress this enough as the ever popular D5 has zero run dry capability. Even with water sloshing through its bearing is experiencing excessive wear! Always keep it flooded and whenever there is low water shut it down pronto.

For best results to prevent dissolved O2 in your coolant, you can boil your distilled/DI water first, allow to cool in completely filled containers and carefully (with as little churning as possible) add to reservoirs. Only when the reservoir is completely filled and the lines flooded should the pump be jogged. Eliminate air gaps as much as possible. The installation of ports/foot valves, drains, etc. allows gravity bleeding before the pump is jogged. This prevents air lock in the system, excessive cavitation and bubbles. In the long run your wetted surfaces will remain oxidation free, and coolant will stay in a desirable state much longer.

The return lines to the pump should be of largest diameter possible especially if there are a lot of bends and elbows. You never should restrict the suction side of a pump - ever! This is especially true in higher power systems using pumps such as the Iwaki RD30, for example.

If you see whirlpools in your reservoirs the flow is too high for them and you should consider another arrangement, parallel reservoirs or reducing pump speed. The worst possible thing you can have is a water fall. It may look "cool" but in a practical sense is actually detrimental to water chemistry and long term cooling performance. (dyes and other "enhancements" also fall into this category!)

If you really must have these kinds of things, set up a dummy circuit that circulates this through these boutique devices, but leave the real functional cooling to a traditional design, keeping the fluid dark as possible with no air gaps, etc.

 

[Aikouka]

Also, microbubbles is something that should never happen in a PC cooling circuit! When filling for the first time, the pump should only be turned on briefly and as soon as you hear its sound change, disconnect power immediately! I cannot stress this enough as the ever popular D5 has zero run dry capability. Even with water sloshing through its bearing is experiencing excessive wear! Always keep it flooded and whenever there is low water shut it down pronto.

Honestly, I just never let the reservoir go dry when I would fill up my loop. I would usually let it go down to around 10-25% before I would kill power and fill it back up to around 75-80%. The only downside to being that cautious is that it will take a few more cycles since you're not pushing as much water into your loop each time.