Monitoring Electrical Impulses of Fish

The tank water is probably quite conductive with the fish pee and poo in it, so you have mixed blessing: there is a fair contact to the fish, but there is also quite a short around them. If you can, go to the tank and measure the resistance between the electrodes.

I'd guess that you cannot get more than some millivolts like a normal EKG, prerry surely less.

If you build an amplifier, prepare it for clumsy handling, including ESD.

It's going to be a differential amp sharing a common Earth ground. The wires won't need any shielding as long as they're kept close to each other. High voltage seems unlikely other than people touching the wires.

The effective signal impedance on those plates might be less than 500 Ohms. Measuring the impedance of water is a tricky thing with all those moving ions and oxide layers to deal with.

What sort of sensors are you using? That's pretty critical. Maybe you could adapt the very hi-z ones typically found on EEC machines for picking up brainwaves?

Yes, that is a given. Wires connected to a fish tank will get lots of rough handling. There is an existing design and the front end is no longer available. I haven't see details on the design yet. I figure the short path will be duplicating the design with newer parts. But I need to figure out the requirements before I can do that. Maybe I can even find an off the shelf replacement. So far I haven't found much at a reasonable price. The one unit is $600 and isn't isolated.

The sensors are just plates in the water, one above and one below the fish. I don't have info on the requirements as yet. I should be able to get more info next week.

Have you thought about using a large coil loop under the tank and capture the RF noise from the pulses ?

Jamie

I don't quite understand your question. (got a link?) Fish do have a lateral line down the sides of their bodies that is used to detect nearby motion.

But I don't think that is what you are asking about.

George H/

No, I'm not asking about the fish detecting voltage. I'm asking about the voltage emitted by the fish. Seems it is pretty low level and will need to be amplified a lot. I am asking if there are other things I need to worry about like a high source impedance which someone pointed out should not be the case because the signal source is shorted by the water resulting in a lower impedance.

A system design exists that was developed by a group and tested over a course of years. Now they wish to get it into production, but for some reason the amplifier board is no longer producible. So a new design is needed. I should have more info on the whole thing this week. I'm trying to understand why this is the only electronics in the system that was custom designed. So far the only unit I've found is a $600 single amplifier. It would be better if I could find a multi-channel unit. I see they say to contact them about that.

Hmm those look like typical opamp type specs... shouldn't be too hard.

One thing with water is that DC is hard to do. Try to measure the DC resistance of water... Then do it with AC. (There may be some minimum frequency.)

George H.

On 11/16/2015 12:32 PM, rickman wrote: ...

"Production" - you mean that there's a market (demand) for a system that measures the voltage emitted by fish?? It must be a pretty limited market.

Bob

Fishermen will pay good money if you tell them where to put their bait. $600 is a lot, but given the specs a good quad or two dual opamps might work. Is there any other reason to know where the fish are?

George H.

Check your electrode millivolts with no fish. Depending on electrode metal, there might be significant artifact caused by surface chemistry, especially low-freq 1/f noise.

EEG/EKG employs low noise electrodes, silver with AgCl. If you're not already using such things, you might want to experiment. I've seen large cheap electrode pads available, also silver coated carbon fiber, etc. is discussed in med magazines.

My only actual experience is from grade-school days, filing down old Canadian nickel coins for DIY ekg and gsr skin electrodes.

The water is going to be a low-z signal source. It should will short out any high-z e-fields or capacitive coupling, unless you're running big transmitters or large CRT displays right near your signal leads.

Just keep the wires close together, twisted pair would be good. Keep away from power transformers, or fluorescent lighting ballasts, which could induce voltage in the tiny loop-area between your pair of connecting wires.

It might be prudent to control the water pH with a suitable buffer. Some easy-to-find materials (pencil lead?) are suitable for prolonged contact with dirty water, won't react... much.