Unity gain op amp circuit design

[guisek01]

Hello, This is my first post here and I'm an amateur with circuit design in general. I'm a neuroscience graduate student and I need to build a unity gain pre amplifier to continue on with one of my experiments. We successfully built a 4 channel board using Texas Instruments' TLC2274AIPW op amp IC. in the PCB design, we had a positive and a negative power trace, each connected to ground via a .1 uF capacitor placed near the IC. Connected to each of the positive inputs was our signal of interest, and the output was connected to the output on the PCB board as well as the negative input for the op amp. the output/negative input was also connected to ground via a 15 MOhm resistor. The design is similar to that seen here at the top of figure 2: **broken link removed**. Note however, that their design differs from ours such that we do not have a resistor between ouput and negative input, and our ouput/negative input is connected to ground via a resistor. Also, we have the power traces connected to ground via two capacitors.

I tried to extend the PCB design so that we could get 8 channels by making the PCB a bit bigger and placing a second IC along with the respective capacitors and resistors. Now I seem to be having a bit of trouble with noise and the stability of the signal passing through in general. I now have 4 power lines (one positive and one negative for each IC), each with a closely placed .1uF capacitor connecting it to ground (there is a single ground plate on the back of the PCB). Could this be the problem? Should I instead have a single capacitor connecting a large positive power trace to ground, and then route the power to each IC using a star configuration, e.g., as discussed here: **broken link removed** ?

Any help on this would be greatly appreciated, as if this works out, I'd like to move up to a 16 channel design and a 4 layer board. Actually, if there are any additional considerations I should be aware of for 4 layer boards, I'd be happy to hear them.

Best
Kevin

 

[keith1200rs]

Am I correct in thinking your non-inverting input is floating? If so, layout and screening of that will be very important. Also, the route for the input to ground will be important.

Keith

 

[guisek01]

Hi Kieth,

I think so, but perhaps I'm wrong. The input is coming from an electrode implanted in a brain on one channel, and a skull screw on another channel. The organism is connected to the ground plate of the PCB (and ultimately earth ground) via a second skull screw. The output of each opamp (which is directly connected to the negative input), is connected to ground via a 15 MOhm resistor. A differential AC amplifier amplifies the difference between the signals from the electrode and the first skull screw that were passed through the opamp. Any specific tops on what I should try to achieve/avoid while designing the PCB?

Best
Kevin

 

[keith1200rs]

This is where I am confused - why put a 15M resistor from the opamp output to ground? I would have expected a large value resistor from your non-inverting input to ground.

I assume you are using screened cable to the electrode?

Keith.