Power Amplifier for EEG Signals

** Do you need the display to be alpha numeric so the plant can reveal it's innermost thoughts and feelings ?

Or would variable luminosity do so it can answer yes/no questions on plant related matters ?

** Think you should consider full audio bandwidth & employing to a voltage to speech converter IC.

I'm sure Analog Device used to make one of them ...

... Phil

Hi Richard,

I can't be definitive on this because I have never done it. If I were you, the very first thing I would do is to esearch what plant physiologists have done in the past - how did they do this? A piece of information is what voltage level have they measured? millivolts? If it's millivolts, you are in luck. If it's microvolts, you will have to be much more careful.

BTW, the 12 volts at 1 amp is the easy part. The challenge for you is the front end, the part that measures the plant voltage. I would think about copying the eeg approaches - using an ionic electrode. Then make a low noise amplifier and put it right on the plant - in situ. That eliminates long wires. If you can elevate it to >>50 mv, then you can do a run to an instrumentation amp with high cmrr on the lines...

I imagine you will have lots of noise - it's almost a dictum that noise is 1/F - means it lurks at low frequencies. People who do these measurements love to move up in frequency if they can. Expect popcorn noise, raspberry noise, bacon frying, and other types that go bump in the dark. The only way I can imagine doing a high frequency chopping is the measure an adjacent, non-active area in a chopper arrangement at 500 Hz, say.

If you can get some AgCl eeg electrodes and a sensitive op-amp, you could get your feet wet.

I remember a book, 'The Secret Life of Plants," by Christopher ______ . But definitely search plant physiologists to see what they have done. And call them up on the phone. See if you can borrow any apparatus from them. No one talks to them.

Sometimes it's better to jump on the heads of knowledgeable experimenters, like a loose cat on an airplane. :)

jb

Both Linear and Analog Devices make a EKG sensor chip. Can't find their part numbers now.

Experimented with both, copared to 'home-made' sensor. The Linear part has less 'fuzz' in the spectrum due to a clever approach. However, does have a spike around 4kHz [from memory] But, you're not interested that high, right?

Recommend search at both their websites for such IC's Get the data sheets and see what is readily possible.

The 12 volts @ 1A isn't that much of a deal however, the input side is. You need to use a quiet front end and what comes to mind at the moment is a chopper amplifier, which is perfect for the job. That at least can bring the low end out of the hole with lot less noise..

Some reading..

Jamie

Our group is attempting to build a low frequency amplifier (DC to

60Hz) capable of amplifying voltage potentials in plants to a level of 12Vpp at 1A.

The idea is to drive a large visual display, based upon this activity, for demonstration purposes. To achieve this effect, there must be no significant delay between the input and output.

What are the best design options for this application?

Richard Clarke

Normal EEG works at a 50 microvolt level and if you try silver chloride electrodes on a plant leaf, it will probably damage the plant. Silver chloride is a old standard, I think they have updated the electrolyte to something else. You could contact places that do sleep studies for sleep apnea testing, they may be able to give you some information.

Shaun

Thanks for the suggestions, but I already have the front end designed.

My OP was asking for advice on a sutiable _power amplifier_ configuration for these DC to 4Hz EEG type signals.

Output 12Vpp at 1A

Some have said that is the "easiest" part of my project.

Just an Op Amp and emitter follower I suppose then.

Richard Clarke

** ROTFL !!! ** Your OP ???

You running some kind of information agency using usenet ??

** You a total idiot or merely a total f****it ??

Or is this a really dumb troll ????

.... Phil

Some ordinary DC coupled audio amplifier should do. This will either require at least +/-8 V dual polarity supply, but if only a single polarity, say 12 V battery is available, you could connect a _floating_ load between the outputs.

While many audio amplifier claim to be DC coupled, they are normally non-inverting amplifiers with a capacitor in series with the other feedback resistor, effectively dropping the voltage gain to +1 at some very low frequency, in order to avoid amplifying any input offset voltages, which could damage the loudspeaker.

If your load is not sensitive to DC bias, then you could short that capacitor and have full voltage gain down to DC.

Having 100 to 120 dB of voltage gain can cause a lot of problems due to feedback, so the inputs and outputs should well isolated from each other. In addition to capacitive feedback from output to input, the required output current of 1 A is relative large, which can also induce some inductive feedback. Good balancing and shielding are required.

The mains 50/60 Hz mains can also cause problems, so a 50/60 Hz notch in the signal chain would also help to avoid external intrusion.

Hello, If the design is ready, has it been prototyped ? Once prototyped, the front-end could be tested and its output checked with an oscilloscope, e.g., after connecting it with a plant. As my supervisor often mentions casually -- "do you know exactly what results to expect". Once the front-end has been checked to work fine, designing and building the rest would be easier. Your statement "Just an Op Amp and emitter follower I suppose then" seems to good to be true.

(Hmm Well Phil, I sometimes get the feeling that you're the only sane person here.) (Where, after all, are the electrical signals in a plant?)

to the OP you can search for power opamps. I use the OPA544T.

George H.

Hello, If the design is ready, has it been prototyped ? Once prototyped, the front-end could be tested and its output checked with an oscilloscope, e.g., after connecting it with a plant. As my supervisor often mentions casually -- "do you know exactly what results to expect". Once the front-end has been checked to work fine, designing and building the rest would be easier. Your statement "Just an Op Amp and emitter follower I suppose then" seems to good to be true.

I would have too agree! You need a well designed circuit to amplify a signal at that low level (50 microvolts for EEG) and get something meaningful out of it and not just amplify noise. You're going to need isolated power supplies, instrumentation amplifiers built with really good components that don't add noise to your very small signal. At 4 Hz and below you might measure galvanic currents made by the battery that you just created by that plant juice and your electrodes. You won't know what you are measuring.

Shaun

this has been widely done - "The Secret Life of Plants" work, 50 years ag, used these measurements. Why reinvent the wheel?