# 1000x Gain Op Amp, battery/5V powered circuit.

#### jwearing

Joined Feb 20, 2007
1
Hi all,

After some expert advice, here's the brief...

Op Amp series with 1000x gain (preferably two op amps).
Inputs and outputs in BNC/Coax connectors.

Boxed and built on PCB.

Now, being a computing science student I'm not 100% savvy on Op Amp tech or circuits. So any help would be appreciated.

I am aware that the infinite gain of Op Amps means that i must simply make the correct resistance ratio over the amps to create 500x gain in each amp, but even this stumps me!!

The batteries also confuse me. Two 9V batteries, one for +ive and one for -ive is my initial idea, however I must ensure that the batteries have an indicator for low power, I have looked at using TRIAC's to ensure that only correct voltage power is passed to the op amp...but another suggestion was a simple low power circuit with LED indicator? Again, i'm non the wiser. I also need the battery circuit to be switched on/off, so I can (in other words) select the power source to be used...

The input/output into the circuit/amps must be coax/BNC (same as the CRO connectors) how can I do this without leaving a signal floating at the output?

Finally, this is meant to be a two channel amplifier. So how can I duplicate this one system and power both at the same time from the same source.

Sorry guys, I know this may seem like a simple system, but it's not Java/C/Unix or MATLAB, so I'm a touch lost!!

Any help appreciated!

James W.
UK.

#### Ron H

Joined Apr 14, 2005
7,049
The gain of series amplifiers is not additive, it's multiplicative. The gain of each amp only needs to be sqrt(1000) if you have 2 in series.
Why are you doing this? Is it homework?

#### Distort10n

Joined Dec 25, 2006
429
You can make it real simple. First gain stage would be 100 and the second gain stage would be 10. Depending on the nature of your source, a non-inverting configuration may be best since the source will 'see' the high input impedance of the amplifier. Ideally infinite, but we live in a practical world. A few giga ohms or higher usually on a typical voltage feedback amplifier.

It is not clear what you want for your supplies, +/-9V using batteries or a single 5V rail. This matters depending on what you expect on the input as well as headroom issues like input common mode voltage and output swing.

If you are connecting a coax cable to the output of the amplifier, you will need to take into consideration the capacitive load this will cause. A large capacitive load on an amplifier can cause undesirable effects like ringing and oscillation. If it is short, it may not be a problem. A quick and easy fix in most applications is a simple series isolation resistor on the output. This decouples the capacitive load caused by the coax from the output and mathematically places a zero in the feedback transfer function which counteracts the pole caused by the capacitive load.

Another issue is the implied dynamic switching of your power supply rails. The issues that you are facing are as follows:
A signal present at the input of the device that could very well be higher than the supply rails for a finite amount of time. A few milliseconds is an eternity in analog electronics. This may forward bias any internal ESD protection diodes and you must current limit to protect them. I am not entirely sure about op-amps, but this could also cause latching of the device and you would have to power cycle to come out of that state.
An invalid output for a finite amount of time. I am thinking the transient could be seen as a really big and nasty ripple and cause high offset. CMRR and PSRR would go to pot. This would actually be cool to see in the lab.
This dynamic switching needs to be addressed. Perhaps the creative use of diodes or logic level FET's to ensure the input is not present at the time of switching?

You can power both amplifiers from the same supply rails. You just need to make sure the supply can meet the current requirements and not current limit when driving multiple devices. I would imagine this would not be an issue. Bypass capacitors on EACH device and on each rail!

#### Ron H

Joined Apr 14, 2005
7,049
Due to the fact that the gain-bandwidth product of any given op amp is a constant, the closed loop bandwidth will be optimized when both amplifiers run at the same gain (sqrt 1000).
Coax loading will not generally cause instability if the source and/or load end is terminated in the characteristic impedance of the coax. Terminating both ends causes a 6dB gain loss, but generally gives better signal fidelity. Of course, your op amp must have sufficient drive capability to drive the load. Many op amps have this capability.