60Hz "click" or clipping noise when differential amplifier is surrounded by salt water

Question

I am a graduate student attempting to take electrophysiology recordings from a lobster under water. In the current set up I am using, two differential amplifier boards are attached the the lobster's back in waterproof casing, with a transmitter and power source in a float being towed around the water surface.

The amplifier boards function properly for bipolar EMG recordings when the lobster isn't in the water, but once the lobster is submerged with the amplifiers on its back, I pick up very large amounts of sinusoidal 60Hz noise as well as a 60Hz "clicking" noise (see attached image). The sinusoidal 60Hz noise I can filter out post recording, but I haven't found an effective way to remove the clicking noise. Both are drowning out the desired physiological signals.

NOTE: the amplifiers have bandwidths of 200-1000Hz, but no 60Hz notch filters. I've run trials with the reference electrode in the salt water, but also bipolar tests with both electrodes placed in the target muscle tissue.

From what I've read I imagine the salt water is amplifying the ambient 60Hz noise, which causes it to increase in the recording once the amplifier is submerged. However, I'm still trying to figure out what about my circuit may be causing the clicking/clipping 60Hz noise. Please let me know if you have encountered noise like this in the past, and if you have any advice on how to address it!

Answer 1

Sound travels very well through water - try isolating the tank sonically or mechanically from whatever you can. This could be as simple as putting a dampening material (carpet!) under the tank. It is a long shot, but I think it warrants an answer rather than a comment nonetheless.

I once wasted a day trying to fault-find a sensitive differential amplifier circuit that was exhibiting almost exactly what your tests are showing: a ~1-2kHz single sinusoidal pulse every 20ms (or the 60Hz equivalent in your case), but no actual mains frequency sinusoid. We tried turning the lights off in the lab, the computers, equipment in neighbouring rooms, everything. We tried laptops, all manner of electrical shielding - we put a grounded tinfoil 'hat' over the circuit board - and even resorted to working by candlelight once it got dark.

When we knew we had lost the battle my tutor rested his head on the desk in defeat, with his ear to the surface, and heard the spurious signal; the desk was vibrating. We determined that the building's cooling/heating system was the source of the excitement (pun intended); it was roughly beneath the lab and the noise could be heard faintly in the corridor outside its room. A heart-warming story.

In the end someone put a scarf underneath the circuit and the interference was gone instantly. Even with the nearby electrical equipment turned back on, mechanical isolation prevented any unwanted signals.

Answer 2

First, realize that the water is in electrical contact with the lobster. Have you grounded the water, or the lobster? Many electrophysiological preps involve a big ground needle electrode, and I think a ground electrode in the water might be an analog here.

If that simple try doesn't work, there are two approaches.

1. find and kill the source of the noise
2. shield the noise.

To shield the noise, first you need to make sure that the noise isn't coming from the tank environment itself. Turn off the pump, heater, chiller, and any other electronics associated with the tank. If the noise is still there, that's good -- it means you can shield it!!

To complete step two, build a Faraday cage. some sticks and some copper screen to encase your setup, with the copper screen going to a beefy ground is the way to go.

You might also peek at http://crawdad.cornell.edu/ and see if you can see any tips.

Lastly, those look a WHOLE LOT like action potentials. I'd move my electrodes to a different part of the lobster, or just hook them up to a dead piece of meat, just to make sure that you're not looking at action potentials that happen to be firing at 60Hz.

Answer 3

As others have said you clearly have some sort of coupling from the 60Hz wall power. The tough part is figuring out the source of the 60Hz and/or how the coupling is occurring.

In a situation like this, certain techniques will really help to narrow down the problem:

1. Set up a stable reproduce case, if at all possible without the presence of the (probably impatient) lobster. That will make it much easier to determine whether a given intervention has an effect on the problem.
2. Make changes to the electrical environment in an attempt to make the problem better or worse. Specifically, making a problem such as this worse can be very helpful in teaching you just what the cause might look like. Turn the lighting system off or on. Temporarily unplug the circulator pump. Dip your hand in the water. Turn the heater (cooler?) on or off. With a minimal set of equipment on a power cord from another room, throw the circuit breakers for the room. Raise the floating power supply out of the water. After any change, note whether the interference gets better or worse, or even just changes waveform. Etc, etc, etc.
3. Keep track of ALL your changes, and the results. If you manage to make things significantly better (OR worse), and you can't remember what you changed, you'll be reeeally annoyed with yourself.