The Gladys Hydrophone

It is time to update my Hydrophone Instructables. My first Hydrophone Instructable is a few years old at this point. Many were built and used around the globe. One of the first things I learned was that there needs to be a ground connection from the recorder to the water. That is facilitated by addition of a ground wire, which caused me to update the original instructable. Now, based on personal experience and feedback from multiple users, I have made more improvements. Enough that it warrants a new Instructable. They cover ease of build, technical improvements and durability.

Changes:

• Input protection circuitry for the Opamp
• Redesigned PCB for ease of wiring and assembly
• The sealing of the Piezo Cylinder for an airspace to improve acoustic response
• Larger mold to provide additional space for the piezo element
• V-groove on the mold to allow easier separation

We also had a sourcing issue with the Piezo Cylinder and Steminc. They are selling off inventory and closing shop. I partnered with JLI Electronics to setup an alternate source.

Quick Review: A hydrophone is a microphone that works in water. Unlike air which is compressible, water is not and carries sound differently. Sound also travels much faster in water. To capture it we need a transducer that picks up pressure variations in the medium and cylindrical Piezo elements are perfect for that. One of the challenges is that they are pretty high impedance electrically. To best capture the signal they produce, we need a high impedance buffer. That is what both of the circuits below provide. See the original Instructables for a more in depth explanation. The second Instructable is for a dual channel version that runs off a 9 volt battery and intended for drop rig use.

Audio Samples:

Toad Fish

Humpback Whales

Icebergs

Microphone Cable: (these have been tested and verified submersible)

Single https://www.redco.com/Mogami-W2549.html

Dual https://www.redco.com/Mogami-W3106.html

XLR Connector: https://www.redco.com/Neutrik-NC3MX-B.html

XLR Boots: https://www.redco.com/Neutrik-BSX.html

Copper Wire: https://www.amazon.com/dp/B0CC569T9J

Silicone Wire: https://www.amazon.com/Fermerry-Silicone-Flexible-Electric-Stranded/dp/B089D13Y1N/

Colorant: https://www.smooth-on.com/products/so-strong/

Resin: https://www.smooth-on.com/products/smooth-cast-65d/ on Amazon: https://www.amazon.com/gp/product/B00E3ZJ9XW/

E6000 Glue: https://www.amazon.com/E6000-237032-Craft-Adhesive-Clear/dp/B004BPHQWU/?th=1

PCB's

P48 version: https://www.jlielectronics.com/diy-accessories/p48-hi-z-piezo-buffer/

9V Dual version: https://www.jlielectronics.com/diy-accessories/dual-hydrophone-pcb/

Piezo Cylinder

JLI-PZ001 https://www.jlielectronics.com/diy-accessories/jli-pz001/

Complete kit (minus the resin, XLR connector and mic cable) here: https://www.jlielectronics.com/diy-microphone-kits/gladys/

Circuit changes:

Well, it turns out the Hydrophone circuit needs input protection. We are using an Opamp, the OPA1642, as a high impedance buffer for the Piezo Cylinder. It turns out if you tap the cylinder it can put out a pretty high voltage spike. Mechanically it resonates around 43 Khz. This means the spike doesn't last long before it swings the opposite way. It can reach pretty high voltage levels. We need to protect our Opamp from those. 

Here is the new input circuitry. This is common to both the Original P48 version and the Dual 9V version for use in a drop rig. 

I added a 1K input resistor in series with the 1Meg resistor. This is a .1% change in overall input level, which is insignificant. Next we come to a pair of Schottky diodes. These do two things different from a normal diode. They conduct much faster, and they start conduction at a lower voltage than a silicon diode. So if the input voltage goes above the positive supply voltage, the top diode will conduct. If it drops below the negative supply voltage, ground in our case, the bottom one will conduct. The .1uF capacitor has a pretty low impedance at 43 Khz. About 37 Ohms. This gives our incoming transient somewhere to go. The cool part is if the circuit is not powered at all, the protection circuit still works and protects the opamp. I had one or two failures when it was not powered on and I think I dropped the hydrophone or banged it on something. More on protecting from that later.

OK, now that we have the protection circuit out of the way. Let's do a quick review of both circuits.

The original P48 version has incoming phantom power on Pin 2 and 3 of the XLR connector which we tap off with a couple resistors. We filter that with a capacitor and then have a zener diode regulate it to 12VDC. Then we have another filter capacitor after a small resistor to get rid of any noise generated by the Zener Diode. Now the not so intuitive part. Op Amps need a dual voltage supply to operate properly. We are going to create one with the two 47K resistors and the 22uF capacitor. This gives us a virtual ground that is about 5.7 volts when referenced to the actual ground. Now onto the signal part. We come in from the Piezo cylinder,  go through the protection circuit previously discussed, and then into the non inverting input of one of the OPA1642 stages that is setup as a unity gain buffer. From there the signal has two paths. First is into a 2.2K resistor and into the second op amp stage. This one is an inverting stage with the gain set by the 2.2K feedback resistor divided by the 2.2K input resistor.  Giving us a gain of negative one. Now we go to Pin 3 of the XLR while the non inverted signal goes to Pin 2. Both signals couple across a 22uF output capacitor and a 47 Ohm resistor, which decouples the opamp. With the exception of the protection circuit this is the same as the original hydrophone.

The dual Hydrophone has two changes on top of the protection circuitry. First is where the 9V battery connects. I have added a reverse polarity protection diode so that if you accidentally touch the 9 volt battery backwards to the connector, nothing bad happens. No one would do that would they?... By the way this circuit can run on about 6.5 volts up to 30 volts. 

After that we are creating a virtual ground just like we did with the single channel hydrophone. Two 47K resistors and the 22uF cap provide that, Now we go to our two identical op amp circuits. Our signal comes in and hits the same protection circuit with the Schottky diodes. Then it goes to a non-inverting gain stage. These have a gain of 1 plus the ratio of the 10K/1K or… Just under 21dB. I found that many small recorders with 3.5mm inputs have a pretty high self noise when operated at high internal gain settings. The OPA1642 can do this with minimal noise added. Far better than my Tascan DR 05X and even my Sony A10. The rest of the output circuitry is similar to the P48 version. We go through a 22uF capacitor to block DC then a 47 Ohm resistor to decouple the opamp. Both signals go to a 3.5mm audio jack ready to plug into a recorder. 

I tested this by banging an unmolded piezo cylinder against my desk while connected to a mic preamp. The preamp gain is as low as it goes and it still pegs the VU meter. This element has a split in it which is why it rings in the audible range. No circuit damage! Success on this front. See the YouTube Video for this in action.

These are both updated with the new changes and available from JLI electronics for purchase.

I made the mold bigger and used Fusion 360 to design it vs TinkerCad the first time. I also added a V groove in the side to ease splitting the mold after casting. What I really need is a reusable silicone mold. Anyone want to take that on? PM if you are willing. I also designed the end caps for the Piezo cylinder to seal it. The bottom one is just a disc. The top has a small slot for the inner ground wire and two small tabs to hold the PCB in place while molding. 

This is the build and testing video. It is time stamped with all the applicable portions

Lets Build them

Here is the new top piece that seals the top of the cylinder and holds the PCB for molding. It slips right between R10 and R11. The top has a slot for the inner wire to pass through. We will seal that with E6000 glue. We are using short pieces of silicon insulated 24 gauge. Most stranded 24 gauge will work. See the parts list for details. Solder one to the exterior silver plating and one to the interior. Inner is ground and outer signal. Let's seal the cylinder. Apply a ring of E6000 to the bottom disc and place the cylinder on top. Let it dry and inspect it. You are looking for 360 degree coverage.  Now do the same with the top seal. This one is a little more tricky as you have to pass the inner wire into the slot.  Once this is done, apply some E6000 to the slot to seal it up. The Cool thing here is that the resin only has a few minutes to leak before it sets up. Once this is dry inspect the cylinder. You are looking for obvious leak paths. Other changes to the PCB are placement of the connections for the cylinder and the diameter of the holes for the wires. This makes assembly so much easier. The Ground and SIgnal come in from opposite sides to be symmetrical. Connect those and then use a small amount of E6000 on the two nubs sticking up and insert the PCB. Let the whole thing dry and then test prior to molding. 

We are using Mogami microphone cable. I get it from Redco and it is available in multiple colors. I have also built pairs of hydrophones using dual microphone cable that looks like Zip Cord. This is also Mogami. These have been tested for over a week submerged.

Single https://www.redco.com/Mogami-W2549.html

Dual https://www.redco.com/Mogami-W3106.html

The first step is to put an XLR connector on one end of the mic cable. Slide the XLR Boot over the Cable. Don't forget this! Now let's tin the XLR insert. Prep the end of the Mic cable so it looks like the photo and then tin the ends of the wires. Shield to Pin-1, Blue to 3 and Clear/white to Pin-2. Write this down or remember it for when you connect the Hydrophone Board.  After inspecting, assemble the connector. Put the strain relief on the cable and slide it up to the insert, then insert the whole thing into the shell. Slide the Boot up and screw it onto the shell assembly. 

For the hydrophone end, strip off the outer insulation. I use a single edge razor blade. Practice this if it is your first time. You want to be able to bend it and have that break the last bit of rubber.  Pull the insulation off and and form the internal wires so they line up as shown. Shield to Pin-1, white to Pin-2 and Blue to Pin-3. Tin the wires and solder to the PCB. Inspect your work! Now let's add the ground wire. You only need one of the hydrophones to have this. I learned this the hard way. Without it, hum and buzz can happen. In one sense this is Rev-3 of the hydrophone project! Rev-2 was the addition of a ground wire. 

Cut a 4-6 inch piece of 16-14 gauge solid copper wire. Form a small loop in the wire. Hook it into the shield wire and solder. Here is the finished product. Connect this to a mic preamp and test BEFORE EMBEDDING!!! 

Another change to the project is the use of a larger mold. This makes it easier to get the cylinder centered and has more height to ensure all the wires are covered with resin. Before molding, lower the assembly into the mold and have it held steady. There are several ways to do this. I recommend a “Third hand” type thing. If you use one with alligator clips, wrap some tape around them to prevent damage to the cable insulation. 

https://www.amazon.com/Magnifier-mlogiroa-Magnifying-Adjustable-Inspection/dp/B08Y1JZHD4/?th=1

It is critical that it not rest upon the edge of the mold.  Once you have it centered and positioned, it is time to mix the resin. The resin has a part A and part B which are mixed together in equal parts and stirred. Wear nitrile gloves when doing this or some other PPE for your hands. Notice the clarity as we first mix it. As the two parts start to react two things happen: It gets clearer and actually feels thinner while stirring. Now is the time to pour. It should just start to get warm. If you keep stirring, it will thicken and you can mess up the molding process. Dont go past 2 minutes when stirring, I know! I messed a pair up while messing around with my cameras filming the build. Onen set pulled some of the color from the green PLA I printed my mold with. Although with other colors, it does not. If you pront the molds yourself use ABS or PLA. PETG is too strong to break the mold apart. 

I had a request for coloring the resin and it turns out they make a dye for it in various colors. Which is perfect for an upcoming project of mine – an immersive hydrophone array! For today,  Let's make some Purple Hydrophones. Following the instructions,  We are adding a couple drops of the tint to Part “B” of the resin. Thoroughly blend this then add an equal amount of “A”. The dye makes it harder to see the two blending so I recommend you practice without the dye to get a feel for proper mixing. Once it gets slightly less viscous, pour it into the mold. Let this setup for at least an hour before breaking the mold away. The reaction causes heat so wait until the whole mold is back to ambient temperature. After this build run, I wanted to make it easier to remove the mold. So I added a V groove in the side of the mold that makes it much easier to split. You still need to score the mold with a blade but it is far easier. Final tips here. Each mold needs 80mm of resin. Ensure the mold is fixed with tape to a solid surface or work bench. Ridgedly support the cable in the mold.  Mold them one at a time. You don't want the resin to start setting up while pouring or even thickening.

I brought all three pairs of audio into Audacity and then increased the gain to normalize things. I noticed no real gain difference between my original design and the ones using an airspace in the piezo cylinder. If you already built a pair you are good. In theory it will help. I plan on using it in all the P48 builds I do.  For the drop rig version I would just skip it. 

I also found that the Aquarians sounded less clear and muddy. Although they do have about 12-13 dB more initial gain. Opening the spectrogram in Audacity showed why: They have almost no signal above 10Khz, While the Gladys hydro’s  easily go  into the ultrasonic range. Here is a screenshot showing this.  

My final final improvement for these is to add some physical protection. A small section of a pool noodle works fantastic as you can slide it one and protect the hydros from impact on dropping or banging. 

This one has the protection circuitry and 20dB of gain built into the circuit. I first tested it in Aruba snorkeling in a small cove near rocks. Then I had a chance to go back to La Jolla Cove. You can hear that at the end of the YouTube Video. 

I am just super impressed with how these perform. They far exceed commercial ones costing many times more. If you are at all interested in underwater recording you will not be disappointed. I plan on building an 8 hydrophone array of these for some immersive audio recordings. Think Humpback whales in Dolby Atmos.