How can I use the Probe Check?Follow
The Probe Check is a powerful tool to give insight into the operation of the instrument. The check gives you “eyes” onto the region where the instrument is making measurements by showing how the signal varies with range. If you are running into a problem with your velocimeter, running a probe check should be the first troubleshooting step. Note that the probe check data is only useful if it is collected while the probe is submerged.
Before presenting different examples, here is a short description of how to run the probe check and where you can find the data afterwards:
Vector: The probe check data is available in two different ways: Before and after each burst a probe check is made, or at the beginning of the measurement when selecting Continuous mode. Continuous probe check can be gathered if the recording to disk function is enabled before starting a probe check in the Vector software (Online > Start Probe Check).
After a bin to ASCII conversion it ends up in the .pck file explained in the .hdr file. ExploreV does not handle the probe check data, so you’ll need to use another program like MATLAB to plot this data. Note that when recording to disk, the data will be in binary mode. The format is described in the System Integrator Manual.
Vectrino and Vectrino+: Click Data Collection > Start Probe Check. When exporting data using the “Data Conversion” function, this data will be automatically produced if it’s available. Look in the directory where your data file is located, and after conversion you will see a few additional files have been created. The instrument configuration information and other (data) file formats are found in the *.hdr file. The probe check data (if it’s contained in the data file) will be in a file with the suffix *.pck. ExploreV does not handle the probe check data, so you’ll need to use another program like MATLAB to plot this data (see bottom of this FAQ for a suggestion on how to do this)
Vectrino Profiler: The Probe Check must be enabled in the Configuration dialog. The probe check data can be found in the “Data” structure when converting the data to MATLAB format (ProbeCheckBeam*).
The figures below show Probe Check displays resulting from different problems. These probe checks are from a Vectrino, but the same effects are in most cases valid for the Vectrino Profiler and the Vector as well. If there are any specific differences, they will be mentioned. The first two images show a probe check presenting the result of what we want to see, from a Vectrino and a Vectrino Profiler respectively.
Vector and Vectrino: Note the amplitude at the beginning of the sampling volume (A), the nice Gaussian peak that indicates the sampling volume (B) and the peak corresponding to the bottom (C). Also note that the signal strength decreases to the noise floor between the measurement volume and the bottom, and that all the colored graphs show roughly the same amplitude and profile shape. The four graphs correspond to each of the four receive beams. The graphs are scaled in distance in mm from the transmit transducer, parallel to the transmit beam. This is also relevant for the side-looking probe. The peak in beam 2 should reach over 100–150 counts. For the Vector, the sampling volume is located at 150 mm from the probe, not 50 mm as for the Vectrino.
Vectrino Profiler: Note that the signal strength is presented in dB and not counts. As above, you should see a broad peak centered around ~5 cm from the profiling volume. Each colored graph corresponds to the each of the four beams. No bottom returns are visible in this case.
Below is a presentation of the most common problems, together with our suggestions on how to improve the result of the check, if it is not related to a damaged probe. For the troubleshooting to be fully effective, a “before” probe check to compare your suspicious probe check against is a good starting point. If you have older data files with good scattering conditions, this should come in handy. If not, use the examples above.
Boundary interference: The sample volume should be centered around 50 mm and the curve should look more Gaussian than the above. Suggestion: Try using a deeper bucket.
Multiple echoes from the bottom reflecting off the surface can also appear in the probe check data. These will typically be at roughly integer multiples of the actual distance to the bottom. Depending on the acoustic path length, an echo closer to the actual boundary position may be the result.
When a side-looking Vectrino probe is placed close to the bottom, the two receivers slanted at 45° may see interference from the bottom. If you see interference in beam 3 and beam 4, try moving the probe upwards a bit in the water column.
Boundary interference: If there are any obstructions in the beam path, the result may look like this.
Little scattering material/low power: The curve showing the sampling volume is “too low” in terms of counts. If the scattering level is sufficient, increase the output power.
Too little scattering material or power level too low.
Bent receiver arm: Please look at the location (range) of the peak in amplitude on the probe check for each receiver. If it is significantly shifted closer or further away, that’s usually an indication of a bent receiver arm. In the figure below, the red and blue curves (representing beam 1 and beam 2) are not aligned with the other two curves. In this case, beam 2 was bent a bit out, beam 1 a bit in.
Check to confirm: Gently place the probe on a flat surface and confirm that all beams touch the surface at the same time. A bent receiver arm will change the calibration since the receiver arm geometry has changed. If there are only minor changes, then you may not notice much difference, but it will introduce a bias to the measurements. Contact Nortek for further guidance.
Bent receiver arms: The curves presenting data for beam 1 and beam 2 are not aligned with the expected curves.
Saturated signal: The probe check should have a nice peak around the sampling volume and not plateau at any point. Try lowering the power level and see whether the result improves.
Saturated signal: The curve flattens at the peak. (The peak around 340 mm from the probe is the bottom.)
Problem with one of the receivers: The red beam is not functioning properly – the curve for this receiver is not aligned with the other three. Check to see whether there is something blocking it, or contact Nortek for further guidance.
Problem with receiver on beam 2.
Problem with the transmitter: Note that this probe is submerged (you will get a similar result when running a probe check in air). The transmitter is not functioning properly. Check to see whether there is something blocking it, or contact Nortek for further guidance.
Problem with transmitter: Either blocked or not functioning.
Grounding (Vectrino: loose grounding wire): The noise floor is varying around 50 counts. Tests in laboratory tanks can sometimes lead to grounding problems that show up as elevated noise levels, but only after the instrument is submerged in water. You will not automatically see the increased noise level in your data if your signal from the water is above the noise, but the increased noise level could look like signal. Suggestion: First check that the grounding wire between the probe and the electronics on the Vectrino is connected (you will need to remove the probe to check this). Second, you may be able to reduce your problems by coiling your cable into a tight bundle and raising the cable above the floor (i.e. placing it on a chair).
Vector: One way to determine the noise level in the tank is to collect with the transmitter out of the water but the receiver arms in the water. The signal you now see is noise only. Grounding problems will disappear when you disconnect the Vector from the computer and from the external power supply. Collecting data to the internal recorder using batteries can be an option to remove possible grounding problems in indoor environments.
Grounding: Keep in mind that grounding problems occur around man-made structures, and are not normally a problem in the field.
Sharp peak around 30 mm (relevant for the Vector): This is just electronic noise that has no influence on the current measurements. There is nothing wrong with the instrument.
Sharp peak in the beginning of the plot.
Scaling: The graph grows beyond the screen. Solution: Right-click inside the graph and select “Scaling” to adjust the range of the Y-axis.
The scaling of the axis can be adjusted in the software.
Test yourself: What is the problem here?
A number of different issues are appearing in the probe check. If you encounter anything similar, try to rule out one potential issue at a time.
Do you want to plot the .pck (probe check data) in Matlab? Try making a code like the following:
plot(d(j,2),d(j,3:6)),grid,title('Probe check data')
Please sign in to leave a comment.