Make simple broadband impulse underwater

Question

I need to make a device (cheaply, so no piezo-based transducers) to generate an impulse underwater in shallow water (<30 m).

It need to do the following:

1. The impulse should be quite broadband (as impulses tend to anyhow)
2. Have a peak pressure level of ~170 dB Lpeak at 1 m.
3. The source should be horizontally omnidirectional (in the far field).
4. Make a highly repeatable impulse (not as important).
5. Only emit sound, i.e. no disposable parts. (2022/08/31 edit)

I'm hoping to use this to measure short-range transmission loss (<1 km) in a quiet area (bg noise <100 dB SPL). If we assume spherical TL then 170 dB becomes 110 dB at 1000 m (20*Log10(1000)=60), leaving at least a 10 dB SNR for me to pick out the impulse at range.

I'm thinking something as simple as a steel ball on an elastic tether that can be stretched to a specified length and then released to slam into a steel cylinder. It does not need to be automatic, but remote activation would be a plus so I can "fire" it mid-depth without getting wet.

Addition 2022/09/14:

I got a dog clicker as suggested ("Rosewood Dog Training Clicker" from "Paradise Pets"). I then estimated the equivalent monopole source level, by a series of clicks at various ranges and measures of the peak received level.

Source level [dB re 1 µPa²] is 146 dB (95% CI range 144-148 dB). This is probably a tad too quite for my use, but a great tool nevertheless (but note spread in data points).

Data points are grey dots, mean is black dots, error bars are ±1 standard deviation, red/blue lines are ± the 95% confidence interval for the true mean (~±2 dB).

I used a calibrated (B&K 4220 pistonphone) Soundtrap 300 HF for the measurements.

Addition 2022/10/06:

Using a regular sized claw hammer and a galvanised steel pole (hollow, ø=3cm, L~2m, wall thickness ~2mm) I got a source level of ~180 dB Lpeak with a drop height of 50 cm. Probably because of problems with consistency of contact patch the level varied considerably:

Might try a steel sleeve, blocked on end, over a steel rod next.

Answer 1

I cannot promise these will meet all your criteria, but they may be worth considering due to their ease of use (if not for this specific case, then for other more generic ones). In the past I have used these as underwater noise-makers:

• Dog Clicker for Training (press thumb to make the click).
• Hammer vertical pole-- have a metal pole/tube vertical in the water and hammer the pole (can hammer in the air).
• Bang Pots. Note: Do not do this in a marina with liveaboards if you want to keep friendly with your neighbors (ahem-- just in case they are napping).

You could consider a modification of the 'banging pots & pans' Where you use a small olympic weight with a line through the center-- the weight is dropped & it sinks to the 'pot' at the bottom.

Perhaps if I think of it-- it would be interesting to make recordings of each of these sounds underwater during a future sea trial just and share the spectrograms of each sound.

Answer 2

Light bulbs have been used in the past to generate impulses underwater. Pretty cheap but perhaps not the most eco-friendly method.

Have a look at the paper from Heard et al. 1997 "Underwater light bulb implosions: a useful acoustic source". They measured characteristics of implosion sounds from different types of light bulbs (source level, peak frequency, implosion depth, etc.). They also propose a couple of simple bulb breaker designs to trigger at specific depths:

Answer 3

I made this rig: The hammer head (1.81 kg, 4lb) kan be lifted to 1 m over the top of the pile, when the whole rig is held by the end of the red string, with the pole in the water. Dropping the weight (hammer head), makes it slide down the strings and hit the pole head. The pole is shorter than I'd like, but I need it to fit in my checked-in baggage.

Impulses are reasonably consistent with a mean Lp of 189.1 dB re 1 µPa² (185.0 to 193.3 dB range from 10th to 90th percentile & 95% confidence interval of ±1.1 dB)

Single impulse exposure level is ~ 152 dB LE

See spread of range-normalised levels in graph below

Derivation of source level: