Nano-amp amplifier - 1nA to 1V

yungman

In MT-059 you posted, look at Fig.1 and equation 1. That explained the noise gain due to C1 with is the input capacitance.

Did you try the copper tape trick to see whether you can get the low capacitance you need?

Wapochief

When I calculate c2 in MT-059, I get a very small capacitance (2*10^-14 f). According to MT-059, if I increase the capacitance, it increases the stability - makes sense since it dampens the gain. So I seem to be stable if I use my 0.13 pf for a linear amplification range.

In order to keep the noise low as described in MT-050, I have to calculate the 1+C1/C2 factor, which is around 1500 in my case. If I multiply it by the noise voltage of the op-amp, (7nV/rootHz) gives 1mV @ 10kHz - which is -59dBv.
This should be the dominating noise factor. Doesn't seem TERRIBLE, just a little high.

yungman

It is all about compromise. Speed you want is quite high so something has to give.

Why don't you do some calculation of using smaller feed back resistor and more gain on the second stage. Sometimes it is not obvious how it will go, you have to calculate it out. You are getting down to the theoretical limit, maybe you should write a quick Excel program and input different resistor value and post gain and compare the noise. I wrote one long time ago and I don't know where it is anymore, it was helpful. Just use the formulas in those MT for calculations.

Also, remember post #23 about two amp solution. If you can split the light, two amps afford you to use smaller PD with lower capacitance. That will really help your speed and noise gain. Not to mention the uncorrelated noise advantage on summing.

Wapochief

Just wanted to let you know the project worked like a charm! Couldn't have done it without you. We ended up using the OPA656, with a 100Mohm and a 0.2 pF capacitor. Ideally the capacitor should have been smaller to get better BW, but I couldn't find any. It was decided a gain of 10^8 was enough. About -90 dBv noise floor, 2 kHz of BW. It was enough. Surface mount worked well, just skipped the flux. Had to shield the board, grounding the shield to a seperate ground to get rid of a large 60 Hz ripple.

Beat out the $$$ amp they were using.

berkeman

Sweet!

yungman

You make 100M surface mount resistor work without flux problem? Sweet. I never thought you can get away with this high value in surface mount!!!! What resistor are you using, long body? Anything special you did?

Thanks for reporting back. People here have a habit of disappearing when they resolve the issue. You never know whether it work or not. When I ask a question and people help me, I always acknowledge what is the final outcome to close the chapter.

Alan

Wapochief

Just an 0805 100mohm surface mount resistor. No flux to avoid lowering the resistance. Didnt even cut out the pcb.

yungman

Now I am learning. How do you solder without flux? This is good information for me. I am doing some music electronics and I am worry about 100K resistors!!! Good to hear this.

Wapochief

Get a small tip, and put the solder on the tip. Hold the resistor down, then touch the solder tip to one end. Should flow across. Used a 1/32" tip at 350 degrees f.

Wapochief

Even a Gohm works if you dont use flux.

yungman

Thanks. What is Gohm?

Wapochief

Gigaohm or 10^9 ohms =P

yungman