Hi guys,
The contribution below is mainly for colleagues who need to operate (not run) a model at a predefined depth without depth rudders; like research sub models.
I am currently digitizing such a 9 kg / 800 mm boat from the late 1980ies / early 1990ies with the goal to quietly “park” it at any preselected depth (down to 1 m).
During this project I had to notice that all small and low cost pressure sensors (e.g. the Honeywell series, see below) are differential-pressure types, meaning that their electric output is proportional to the difference of the internal hull pressure and the actual water pressure.
The sketch below (cross section) is illustrating the internal structure of these sensors
So far so good.
In principle one could use these sensors as they are, but when it comes to precise depth measurements, things become difficult.
Almost all of our piston- or bladder-dive tanks, as long as they are located within the sub’s tight hull will change the internal air pressure:
When these units pick up water (flooding) they will expand or a piston within a dive tank will return to its outer position.
Result: The internal air pressure will increase.
Example:
Looks harmless, but is not at all !
In water, 1 mb pressure is corresponding to 1 cm depth; meaning that these differential-pressure sensors will indicate a depth 50 cm less than real.
Absolutely not tolerable ! !
Reason: These commercial differential-pressure sensors have a tiny hole at their base plate to pick up the ambient pressure (figure above). This port can easily be sealed by a piece of clear tape. By these means, the pressure sensor will NEARLY, but not completely be independent from pressure impacts.
However, we need to face another fact; especially of importance during warmer or even hot summer days:
Example:
Result: Depth sensor data are now exceeding the real depth by 37 cm.
Even a taped sensor will experience this pressure change, since its small internal reference volume (figure above) will also be affected by this internal temperature change.
WAY OUT: TOTAL air pressure sensors, such as the MS583x series; they are also easily available from the NET.
As an example, the MS5837 is covering a pressure range of 30 bar (30.000 mb).
Agreed, 30 bar looks shocking for our applications, but these sensors are equipped with an internal 24 bit ADC, meaning that their output signal is resolved in 16.777.216 steps. A little drawback is the fact that these units are delivering digitized data via the I2C protocol and no simple analogue information.
This is not regarded as a drawback, since a sub with precise absolute depth keeping capabilities will certainly be equipped with a microcontroller.
These tiny sensors achieve a reproducible depth resolution of 1,8 mm (yes: Millimeters) and ARE TOTALLY INDEPENDENT FROM ALL ABOVE TEMPERATURE IMPACTS.
As mentioned above, this is mainly for subs that need to be operated at a constant depth – preferably without depth rudder.
As usual: Comments and suggestions welcome !
Cheers
Volker
The contribution below is mainly for colleagues who need to operate (not run) a model at a predefined depth without depth rudders; like research sub models.
I am currently digitizing such a 9 kg / 800 mm boat from the late 1980ies / early 1990ies with the goal to quietly “park” it at any preselected depth (down to 1 m).
During this project I had to notice that all small and low cost pressure sensors (e.g. the Honeywell series, see below) are differential-pressure types, meaning that their electric output is proportional to the difference of the internal hull pressure and the actual water pressure.
The sketch below (cross section) is illustrating the internal structure of these sensors
So far so good.
In principle one could use these sensors as they are, but when it comes to precise depth measurements, things become difficult.
Almost all of our piston- or bladder-dive tanks, as long as they are located within the sub’s tight hull will change the internal air pressure:
When these units pick up water (flooding) they will expand or a piston within a dive tank will return to its outer position.
Result: The internal air pressure will increase.
Example:
- A boat’s inner hull has a volume of 10 litres (10.000 ccm) and the dive tank / bladder has a volume of 500 ccm; the pressure difference between “surfacing” and “flooding” condition will be 1,05.
- Assuming a theoretical ambient air pressure of 1000 mb (prior to launch), the internal pressure will increase to 1050 mb.
Looks harmless, but is not at all !
In water, 1 mb pressure is corresponding to 1 cm depth; meaning that these differential-pressure sensors will indicate a depth 50 cm less than real.
Absolutely not tolerable ! !
Reason: These commercial differential-pressure sensors have a tiny hole at their base plate to pick up the ambient pressure (figure above). This port can easily be sealed by a piece of clear tape. By these means, the pressure sensor will NEARLY, but not completely be independent from pressure impacts.
However, we need to face another fact; especially of importance during warmer or even hot summer days:
- Prior to launch the sub will be at ambient air temperature.
- The difference to water temperature can easily exceed 10°.
- After launch, the sub’s hull and a few minutes later the internal air temperature will be decreased from former outside temperature to the lower water temperature.
- According to the “General Gas Law”, pressure is directly proportional to temperature, meaning that the internal pressure will get lower, correspondingly.
Example:
- Outside temperature (summer condition): 30° C or 303 Kelvin, correspondingly
- Air pressure: 1000 mb
- Water temperature: 20°C or 293 Kelvin, correspondingly
- Temperature inside the hull after a few minutes: 293 / 303 = 0,967.
- Pressure inside the hull a few minutes after launch (originally 1000 mb): 967 mbar
- Pressure decrease: 37 mb
Result: Depth sensor data are now exceeding the real depth by 37 cm.
Even a taped sensor will experience this pressure change, since its small internal reference volume (figure above) will also be affected by this internal temperature change.
WAY OUT: TOTAL air pressure sensors, such as the MS583x series; they are also easily available from the NET.
As an example, the MS5837 is covering a pressure range of 30 bar (30.000 mb).
Agreed, 30 bar looks shocking for our applications, but these sensors are equipped with an internal 24 bit ADC, meaning that their output signal is resolved in 16.777.216 steps. A little drawback is the fact that these units are delivering digitized data via the I2C protocol and no simple analogue information.
This is not regarded as a drawback, since a sub with precise absolute depth keeping capabilities will certainly be equipped with a microcontroller.
These tiny sensors achieve a reproducible depth resolution of 1,8 mm (yes: Millimeters) and ARE TOTALLY INDEPENDENT FROM ALL ABOVE TEMPERATURE IMPACTS.
As mentioned above, this is mainly for subs that need to be operated at a constant depth – preferably without depth rudder.
As usual: Comments and suggestions welcome !
Cheers
Volker
Comment