Best way to do a virtual ground?

[large_ghostman]

No; please tell us.

ak

Nothing much to say, besides it would be better if i get something set up and answer some of those questions above. I dont have a schematic just a vague description, but i have a good idea what was intended. I have also had a email back from the guy who built the first one and wrote part of the paper. I dont think copying what he did is much of an option, they cant get parts like we can, so maybe if i try and sort out what was intended and take it from there.

The main problem i have is it needs altering as it dosnt work properly as it is currently, i also have to keep in mind he is going to try and copy the modified version. I am not sure where to start with this!! My 12V supply is a decent HP power supply with good regulation and current sensing etc, they have a computer power supply. I should probaly copy that other wise they wont stand a chance of redoing it.

On the 0.8V rail how much current am i likely to be able to get? What we are after ideally is a fairly strong alternating field, i wont know how strong until i try it and see what it takes to line the cells up or disrupt the membrane. Only thing i can think of is to use a microscopy dye and watch for it to migrate into the cell (or not!).

Opamp choice wont be large, but i do know they have some scrap equipment that might have some parts they can use. I better not mention where they are, but its a miracle they got anything to work with.
Give me a few hours and i will be back with as many answers as i can get. If it makes sense i think the positive and negative cycle are meant to be different amplitudes, its hard to tell as the paper dosnt translate that well.

The idea of a virtual ground was to do away with a dual psu, all they have is a old 150W computer power supply thats pretty old. How much would it help if i could drop the current requirement? This might be an option if i use a smaller diameter tube.

 

[large_ghostman]

I am going to see if they can get hold of a laptop charger (two pin only type no earth connection), some are ~30V and might give us more options. I could go my own way on this, but the other person gave me the idea and did the original work. Its not there fault they find themselves in a horrific situation.

I am amazed they get any science done at the moment, worse still getting caught doing so would not end well.

 

[unclejed613]

if you need a low impedance rail splitter with the ability to handle several amperes through the virtual ground, this amplifier (connected as a unity gain amplifier with a voltage divider feeding the input) should work. it's the same as using an op amp rail splitter, but is capable of much more current. the ground impedance is 5 milliohms. the offset of the ground is 14mV, which can be trimmed by placing a potentiometer between R1 and R6.

 

[large_ghostman]

Thx for that, i am having a hard time contacting them at the moment. But i can get started my end .

 

[unclejed613]

that was going to be one of my "When Was This" circuits from about 1970. i've had to use high current rail splitters from time to time. an interesting variation of the rail splitter is the QSC amplifier. the power supply rails float, and the amplifier "output" is ground, and the actual speaker output is taken from the power supply center tap. there are a couple of advantages to such a design. a) the input stage can be an op amp running off of +/-12V, and b) the output stage (operated common emitter rather than common collector) can have the collector heat sink tabs of the output devices bolted to the heat sink without heat sink insulators.

if you've never seen an amp like that before it can throw you for a loop seeing audio on what appears to be the power supply rails.

 

[large_ghostman]

Interesting circuit, nothing at all like i had in mind. Pretty clever actually, any chance i can have the sim? worse case scenario i might need to change the transistor there end depending on what they can get. At the moment they dont even have a classroom left! 18 months we been working on this thing and the last six months has been hell over there.

 

[unclejed613]

ok, here you go. if you are using the latest version of LTSpice, you need to find the MJE2955 transistor model from an older version of LTSpice, and add it to standard.bjt i might as well include a well stocked standard.bjt file you can copy paste from.

 

[large_ghostman]

Thx alot, I like that circuit alot. I think we have the 0.8V sorted out!! Turns out we were over thinking it a bit! Ok alot

 

[ronsimpson]

I changed the transistors to 3A.
Emitter resistors could be removed.
R2 is used for light loads. Power from the op-amp goes to the load with out the transistors if the voltage across R2 is less than 0.6V. I would choose a amp with some drive current. Pick R2 so at about 1/2 of the amps max current you get 0.7V across R2. (transistors just turning on)

 

[large_ghostman]

Thx Ron, I will need to mess with the low voltage one, i got a chance to log onto sci finder today. Normally just about every paper ever written pops up on it! Its really expensive so our uni restricts blind searches, but even so i have had a hard time finding the info i need. So is likely to be a try and see joby! I can start to post alot of this up as much of it isnt going to be used in the final research. This is more research to do the research if you see what i mean .

Really wish i could get a proper camera or the correct adapter for my camera! But my scope is a slightly non standard size, then i could show you this stuff in action!