Vector ping rate

I understand the Vector transmits pings at up to 250Hz.
Can this ping rate be controlled ?
How many pings are transmitted per measurement (ie data record) if the Vector is sampling at 4Hz (or 16Hz - or at any sampling rate) ?

Hi

For a given velocity range, the internal ping rate is constant and independent of the "sampling rate" used in the software setup and which tells how often the Vector should output data.  In order words, a 4 Hz sample averages twice as many pings as a 8 Hz sample.

The base internal ping is different for each of the velocity ranges:

Velcotiy range:     Internal ping rate (all three beams):
0.01 m/s                70 Hz  
0.1 m/s                250 Hz
0.3 m/s                250 Hz
1 m/s                  125 Hz
2 m/s                  125 Hz
4 m/s                  125 Hz
7 m/s                  125 Hz

At least, this is what we think we are doing  " />'>

There is a feature in the Vector to save the power consumption by lowering the internal ping rate but it has not been implemented in software because most scientists want to optimize for performance.

- Atle Lohrmann

Hi,

If I set the ADV for 1 m per sec velocity range, I see that it pings at 125 Hz. Then I specify what sampling rate I desire as well. I was wondering how the data are used then to arrive at the final sampling rate. For example, I sample at 64 Hz then is the resulting data the average of two samples or if I sample at 4 Hz, is the data just the average of 31 samples..... If so then is 4 Hz the same as averaging 16 of the 64 Hz samples or are there SNR benefits to the 4 Hz data. I believe that Volgaris and Trowbridge (xxxx) discuss this vaguely but I did not understand it. Plus I suspect that Nortek does some proprietary things here to reduce the noise (sort of indicated this is possible in some signal processing literature I read recently.)

Thanks

Dave

Hi Dave,

Each Vector measurement interval uses all of the available pings. So, if the underlying ping rate is 125 Hz, and the instrument is set to output data at 64 Hz, then most of the output records will be averages of 2 pings, but a few will have only one ping. At 4 Hz, most of the records will hold an average of 31 pings, but about 1/4 will average 32 pings.

-Lee Gordon

Thanks Lee,

Just to clarify this. If I sample at 64 Hz and then average them in bins of roughly 16 I will get the same data value that I would have obtained with 4 Hz data...

Dave

Hi Dave,

Exactly. Average 64 Hz data down to 4 Hz, and the single-sample uncertainty is the same as if you sampled at 4 Hz to start with.

An interesting corollary is that the noise level (noise spectrum) is the same at all sample rates. Another interesting corollary is that sampling this way effectively suppresses aliasing.

-Lee

Hi

Just to be clear: The two averaging processes are not mathematically identical so you will not get exactly the same answer if you average for 250 ms in the Vector (4 Hz) or if you block average four 62.5 ms samples (16 Hz).

In most cases, the difference is negligible.  However, for data with low SNR we recommend averaging internally for as long as possible since the internal averaging tends to suppress data points with low SNR (i.e. give them less weight).

- Atle Lohrmann

Hello,

I was debating with my committee on how the pulses are transmitted and we are a bit confused.

Case 1. The emmitted pulses are formed at a constant frequency (say 125 Hz) and the pulse-pair processing is then accomplished on any pulse return and the next consecuitive return. Thus there is a constant time interval between all emmitted pulses. (Easier to manufacture, I think)

|                                                           \
|-----o-----o-----o-----X
|                                                            /


Case 2. The emmitted pulses are generated in pairs that are at a shorter time interval than 1/125 seconds and the separation to the next pulse pair is 1/125 seconds. These two pulses close together make up the pulse pair for the signal processing and phase calculation.

|                                                             \
|----o-o---o-o---o-o---o-X
|                                                             /

I think case 1. but the bulk of the others seem to think it is case 2. Can you please help to end this debate?

Dave

Hi,

The pulses are transmitted in pairs as described in case 2. The separation between the two pulses in a pair define the velocity range. The next pulse pair is transmitted some time after both pulses in the previous pair have been received.

- Sven Nylund

Hi, Then to clarify, does each sample of the 125 Hz rate for a 1 m/s velocity range correspond to a single pulse pair rather than a singe pulse? I am just trying to understand the separation between pairs as one should expect for a given velocity range. Thanks, Dave

Yes, each sample of the 125 Hz rate corresponds to a single pulse pair. Strictly speaking, the internal ping rate and the internal sample rate are not the same. From a data quality perspective, the interesting parameter is the internal sample rate, which is shown in the table for the various velocity ranges.

- Sven

Hi!  Two follow-up questions

(1) Is there a table like the one above for the Vectrino?  

(2) If the sample rate determines how many pings are ensemble averaged into a single reported velocity, then what role does the Sample Volume Height play in this?  The Vectrino manual implies that the Sample Volume Height affects the statistics of the ensemble average.

Hi there

The number of the samples is proportional to the size of the sampling volume.  In other words, sampling in time and space is equivalent when it comes to the statistical properties of the velocity estimates.

I will try to make an equivalent Vectrino table tomorrow.

Best regards, Atle Lohrmann

Vectrino ping rates:

Velocity range   Plus(200Hz)    Standard (25Hz)
0.03m/s            426                107
0.10m/s            667                167
0.30m/s            1124              281
1.00m/s            1754              439
2.50m/s            1818              455
4.00m/s            2564              641

Hi there,

Is there any table which exists or that you could provide which specifies the time lag between the first and second pulse of each pulse pair?
I am trying to explain the position of the weak spots in my measurements and would find this very useful.

Sam Harding

 

Previously martin wrote:

Hi,

The pulses are transmitted in pairs as described in case 2. The separation between the two pulses in a pair define the velocity range. The next pulse pair is transmitted some time after both pulses in the previous pair have been received.

- Sven Nylund

 

Here's a table of the approximate weak spot locations (they'll vary a little bit because of speed of sound): http://www.nortekusa.com/en/knowledge-center/forum/velocimeters/30180961