Low cost sonar

Well done on developing the thrusters. Please can you turn your attention to a low cost underwater sonar. It is really necessary!

1 Like

Stupid question … what type of SONAR … are you looking for a passive system, an active system … a mix of both. There is a TON of stuff involved in processing and analyzing SONAR data and I can tell you now, for something that actually works there is no “low cost” in the name tag. Then again, it depends on what you consider low cost to be.


Can’t speak for Ed, but I’m going to guess (unless he’s planning a remake of “Hunt for Red October”) he’s looking for an active sonar. :wink:

Harold’s right, sonar for ROV’s aren’t cheap. A Tritech Micron (sector scanning sonar) will run about $15,000 and a BlueView P90 (imaging sonar) will run you about $50,000. Imaginex also makes a sector scanning sonar that costs a bit less than the Micron, but not by much.

There are a couple of consumer sonars that might be worth looking at. Humminbird offers a “360 Scan” sonar which paints a 360 degree scanning image. Range and resolution aren’t that great but cost is around $2000-3000. Another is Garmin’s Panoptix, which is a forward scanning sonar and runs around the same cost. I don’t know of anyone who has tried repurposing either of these for ROV work, but a couple of us have talked about it.

One of the things I see is no matter what you have for a sonar system then you need to seriously think about an guidance system because you can ping all day long but unless your platform knows exactly where it is the data is relative to it … and you are once again playing a guessing game.

You need to get a good navigation fix on the surface and run dead reckoning below the water.

Next on the list is acoustic layers. There are various layer types in the near surface that will screw your day up. You can have surface ducting which can extend to almost 100 feet below the surface, then you get another layer that doesn’t go to deep … then after you get deep enough it is pretty much a increasing sound velocity profile with depth.

You might be dealing with salinity conditions depending on how close to shore you are etc.

This is not a subject matter that doesn’t have a lot of issues surrounding it and with each issue beckons money.


Just looked at the specs on the Tritech Micron … you should be able to have awesome resolution at the transmitting frequency (700KHz) but it is going to get the crap pounded out of it by anything floating in the water. They say that system has a range of 75 meters … I don’t know if that is their range gate for it or it the absorption at that extremely high frequency is what is killing the range.

Look at the stuff that BlueView has … NICE but again the range is going to be VERY short.



@Harold - I’m ex Navy too (in fact former ASWO). I think you need to think on a different scale when talking mini-ROVs. Most never operate in water deeper than 100 feet and with a tether that’s only 100 - 200 foot long, 30 meter range on a sonar is usually good enough. If there were a standard sonar for mini ROVs it would be the Micron and the BlueView is for people with deep pockets (ie. government, SAR, etc.).

There’s a fellow in Louisiana who apparently is testing the Humminbird 360 scan. So far no word on that project.

The Garmin Panoptix is something I’m looking very closely at. The entire sonar system is contained in a potted and water tight transducer pod. Two cables feed 12 vdc and ethernet. It’s designed to operate in 10 feet of water, but there’s a chance it may be able to handle 50 ft or even deeper depths. If it will, then that makes it a possibility for a lot of shallow water projects. I’m working with a ROV manufacturer who is currently investigating this but so far don’t have enough info to say whether it will work or not. Fingers crossed it will.

There was also a Russian ROV manufacturer who was offering a Micron-clone for around $1000. When I last spoke with him (a couple years ago) he had no interest in exporting to the US. Too bad, it might have been a big success here.

Paul, yeah … guess there is a small difference between an old BQQ5 sonar and what is going to be used on an ROV.

The transducers are not that hideous to build it is the software that is a big deal. Pinging and getting a return is a piece of cake … doing something with the data sucks. I can make pressure proof canisters all day long that will handle further than the ROV will and I have molded acoustic windows for transducers before. The biggest issue is impedance matching of your materials so you basically have a acoustically transparent window for your transducer.

I was running some quick calculations on the transmitting frequency and these boys are getting measurement resolutions less than a 16th of an inch! That is pretty darn good.


@Harold - Would it be possible to put the Garmin transducer in a greater depth rated housing without reducing its power too much? And if so, any clever / inexpensive ways to do it? Here’s the transducer I’m talking about: https://buy.garmin.com/en-US/US/shop-by-accessories/trolling-motor-mount-transducers/panoptix-ps31-forward-transducer/prod503830.html#gallery-dialog

I looked at the ones you had posted in an early message and was checking this one out too. What is with the prop on the back of it? Or is it the box above the prop then … assuming that is a trolling motor? Now if you gut their housing … yes it would be possible to transplant it to another housing and make an acoustic window for it that was pressure proof.

I wish there were more pictures available, more so some real good closeups. Know anyone who has one that can take pictures?

Wonder why they didn’t a single cable. I guess they are slapping that thing together for cheap at least on the transducer head.

What were you thinking of for a data cable etc. to the ROV? Ethernet would be no big deal and the power could actually come off the ROV power packs. Another cool thing would be to put it on a gimbal mount etc. so you could steer it around.

All, thanks for the response. I was really talking about low cost hobby mini-ROV’s here.

Paul has the right idea “@Harold – I’m ex Navy too (in fact former ASWO). I think you need to think on a different scale when talking mini-ROVs. Most never operate in water deeper than 100 feet and with a tether that’s only 100 – 200 foot long, 30 meter range on a sonar is usually good enough.”.

In above-surface hobby robotics we have a huge range of low cost ultrasonic sensors i.e. Maxbotix, PING, SRxx series and so on.

These ultrasonic sensors are easily interfaced to microprocessors, and various robot navigation and scanning schemes can be programmed. There are many examples and tutorials on the web.

The sensor concept is there - someone just has to modify the sensor housing and ultrasonic frequency for underwater use!



Ed, is there already software developed for the items that you have described? What type of range are you looking at for the sensor to be able to look at?

Another thing that could be adapted is when the ROV is close to an object, if you jack up the frequency, which will operate better if you have multiple transducers, you could range gate by increasing your pulse rep rate to generate a faster updating picture.

I guess the easiest thing would be if you could gather some data for what is an air operated system, which is what I am assuming you are talking about, and then it would be a matter of adapting a housing and creating the proper acoustic window etc.

If you want to go that route, start gathering. I have limited time but I don’t mind tossing in help on various projects.


The best photos I’ve found so far are here: http://www.thehulltruth.com/marine-electronics-forum/650637-garmin-panoptix-3d-forward-scan-7.html#b

The transducer is roughly 3" x 6" x 1" thick. I suspect Garmin used 2 cables to be consistent with their other sensors. This works well for me as I plan on supplying the transducer 12 vdc from a battery on the ROV and then send the signal up topside on a separate “ethernet” cable (or possibly using a few spare twisted pairs in my ROV’s tether).


Here is an example of a low cost above ground ultrasonic sensor;


The range of these devices in air is usually 3m - 10m. If this sensor could be modified to work underwater it would be perfect.

The Youtube video clip in the middle of the webpage gives some examples on how the ultrasonic sensor can be incorporated into a small microprocessor robot.

The software is written for the robot’s microprocessor controller - it can be Arduino, Propeller or any microprocessor. It is not difficult and there are many pre-coded modules, tutorials and examples available.

By the way, these Parallax users gave you a really good recommendation on their forums - that is how I got here.


Someone gave me as a recommendation??? Weird … I thought the Post Office was the only ones with my picture :slight_smile:

The thing with those little ranging ultrasonic sensors is the power output is pretty low. Now granted, water is going to couple the pulse a lot better than air because of it’s density but again, what exactly are you hunting for specification wise?

The problem with any transducer is it is only one beam of data. If you don’t steer it then it has only the axis that it is transmitting on. That is what I am trying to figure out is what do you want. Do you want some sort of up close underwater range finder … like a little distance to target thing or what Paul and I have been talking about something that can give you enough data to actually create a “sound” picture of what you are looking at with pretty good resolution.


@Ed - I don’t know if you’re on Facebook, but if you are you may have seen a post by one of the OpenROV developers who is using sensors like you posted for a hydrophone. It looked like they potted two pairs of sensors in an acrylic tube. There have also been a couple of other folk who have tried to build their own sonars, but so far I haven’t seen any successfully completed.

Now that we have OpenROV and BlueRobotics thrusters, can a sonar based on some open architecture be far off? Perhaps something based on a cheap transducer and open source software? Wish I had the skills to develop such a thing. I think the demand is there.

Paul, those are some interesting photo’s but no smoking gun :frowning: I see in places that they are talking about the transducer hanging in there are 40 mph in the water. That speed would translate to a lot of pressure that the device is filling. Stick your hand out the car window and 40 miles an hour with the flat of your palm facing forward now translate that to the density of water. I have a feeling that the transducer might go down a good distance just the way that it is.

Now what would be nice to see is the back side where the cables come in at. The front and side areas that I can see are totally seamless. I am thinking that it is possibly a potted Shore D polyurethane such as EN-9. Looks like that stuff and it has been used for various transducers for potting material. It is Black in color and is hard as a bowling ball when cured. Seeing the back side would show if the unit has screws or anything else. It might be totally encapsulated and I would guess that it will stand a good deal of pressure.

In addition, the cables coming from it have to withstand the type of hydrostatic pressures that the unit is too while they are both going at the same speed. I would venture to say that the depth for this unit could be deep as you could stick a cable in the water.

Now one thing you can do if the unit can withstand pressure to 100-200 feet is splice mold an extension to the cable. One of the cheapest kits that you can buy that I know works without a doubt is a product from 3M.

Look up the 3m 82-F1 splice kits. We use those on submarines for splicing cables in the ballast tanks. We helped with a life cycle study to qualify a splice kit for outboard use and a bunch of companies kits were tried. The testing included accelerated aging in seawater along with 40 hour runs at 1000 psi between aging runs until we got through what was considered a full seven years worth of aging etc. The 3M 82-F1’s were the only ones that survived and were qualified for use by the Government for US Navy submarines.

The kits are not cheap but one splice kit could do both cables at once. Like I said, I know these things work because I helped test them about 12-15 years ago and we still use them.


Attached is a photo (or more likely a 3D rendering) I found of the flip side

You made a good point about the water pressure at 40 mph, which might be really good news for what I want to do. I’m familiar with the 3M splicing kits, but in this case I plan on mounting the transducer at the base of ROV and then bring both cables to the electronics housing through cable glands. Then from there (hopefully) connect the ethernet to some spare twisted pairs in the tether. If that won’t work, I’ll probably just splice in come Cat5e to the transducer’s cable with a splicing kit.


I suggest RS485 as a communication protocol. It has good range, and is much easier to implement than Ethernet … assuming that people mean the protocol rather than just using UTP wire.

I am interested in this thread for two prospective purposes.

  1. Search and rescue. At some stage, I would like to build a ship mounted system with long-range narrow beam, horizontal scanning. What sort of power and frequency would allow human body detection at 1km, or is that even feasible ?
  2. ROV for a very specific salvage operation in about 30m of water. I would want a camera and some manipulation to attach a hook or 2. This operation, if it ever happens, would be off the South coast of Wales.

I’ve begun testing the Garmin Panoptix multi-beam sonar. My initial report is posted on my company’s website here: http://marinesimulation.com/?page_id=1517

@Paul cool stuff! Are you planning on putting the SONAR transmitter on a pedestal so you can rotate it etc? That might be a neat option. If you are hovering in the water and you can rotate the head 360 degrees to give you an idea of where to steer next.

Just curious, have you tried to snoop the communication traffic to determine any type of data package from the head? Was curious about that because if that can be cracked, then you can upgrade the software to do anything you want.