TI E2E Community
Precision Amplifiers Forum
High Gain Amplifier Design
It is intended to design an array of amplifiers with a gain of 1200 on each channel.Input signals are between 1uV to 3uV with a BW of 100Hz to 100kHz.I intend to use TI op-amps.I have multichannel amplifiers but they are noisy.Your feedback and tips are greatly appreciated.
In order to obtain a gain of 1200 and bandwidth of 100kHz with a single amplifier, it would be necessary to use an op amp with a unity-gain bandwidth greater than 120 MHz (1200*100kHz) and this will typically require IQ>10mA per op amp. An alternative would be to obtain the required gain with two stages, for example G1=30 and G2=40, in which case op amps with just 5MHz unity-gain bandwidth would be sufficient (40*100kHz=4MHz) - this would dramatically lower the total current requirement especially in an array amplifier design and at the same time greatly increase the number of available amplifiers to choose from.
It is not clear from your post, however, what are the other op amp requirements like power supply voltage range, noise, offset, drift, input/output ranges, etc. and without these we are not able to recommend to you any specific op amp model. If you need further assistance, it would also be very helpful if you could provide us with the actual schematic of your design.
Marek LisSr Application EngineerPrecision Analog - Burr-Brown ProductsTexas Instruments - Tucson
It is always desirable to have an op amp with ideal characteristics like zero offset, drift, and no noise but when we ask about the required op amp performance we mean by this what are the minimum (worst) acceptable specs of the critical parameters; asking for zero-drift amplifiers without it being absolutely necessary, dramatically narrows down the selection list of the available amplifiers and increases their cost.
Also, how do you define the required noise of less than 1uV? - is this 1/f noise for the frequency range from 0.1Hz to 10Hz or is it a broadband integrated noise which in the case of BW=5MHZ in the gain of 30 would required ~2nV/rt-Hz broadband performance? Vn = 2[nV/rt-Hz] * (5e6*1.57/30)^0.5 =~1uVrms
Since you have not provided us with the requested schematic of your application, all I can do is to provide you with the TI search tool - see below:
Thank you for your great support on this project! I am trying my best to explain our project.There is no schematic or block diagram prepared yet. But I can give you an explanation.Input signals are carried via electrodes that are fixed on "Glass". Signals are then given to a "Precision Amplifier" for amplification.The output of amplifier is then fed into ADC and then to a dedicated PC.
Input signals are spike signals of 5uV - 10mV that last between 10uS and 10mS.Our power supply is USB. (A regulated 3V is obtained from this source)
Your question: "How did I define the required noise of less than 1uV?"Answer: We have used a multi channel amplifier array Eval Board and have measured noise of 1uV without connecting input signals.
Now, I have decided to do the amplification in two stages for a total amplification of 1000.Please let me know what parameters you need from me apart from the ones given above.
Do you still need information on "Offset", "Drift", and "Noise"?
Please let me know of an Application Note that relates to my project.
Please see my comments below:
The input signal voltage of up to 10mV is an AC signal but what is the required DC input common-mode voltage with respect to 0 to 3V supply? Depending on the configuration used, a different input voltage range may be needed - therefore, you need to decide on the circuit configuration before you may begin searching for op amp meeting the circuit configuration needs. For example, if you use an inverting configuration with the non-inverting input voltage at mid-supply (Vcm=1.5V), any op amp with 1.5V or better input common-mode voltage range to the rail will do the job. However, if the input common-mode voltage is biased at GND, a rail-to-rail input op amp will be required.
When it comes to the output range, since you plan on driving ADS, most likely you would like to have an op amp with rail-to-rail output range. However, no op amp may swing all the way to either of its rails, thus the question becomes what is acceptable? 500mV or 10mV from the rail? Thus, using USB port 3V single supply, the output range could be .5V to 2.5V, or 10mV to 2.990V, respectively.
It seems like you have determined the noise level of your board but what op amp have you used for the measurement? Are you trying to find a replacement for for existing part and if so what is the part? I am not sure I understand this - you have an evaluation board but you do not have a schematic of the board to share with us?
We will either need to see the schematic of your board to help you figure out what the limitations are, or you need to determine the required specifications and tell us what they are so we can recommend a part meeting these specs - there is no way around this.
We have used RHA2116 from Intan Technologies which has a fixed gain of 200.
You have asked, "Are you trying to find a replacement for for existing part and if so what is the part?" and here is my answer: Yes, we are. Will INA163 work for our application? If it does, we are happy to move forward with the design because most of electrical characteristics of INA163 are very good for our design. Or would like to recommend another op-amp?
As I've mentioned before, we have decided to go for two stages of amplifications, each with 31.6dB gain.In that case, the op-amp of a GBW of 5MHz. (38.1 * 100kHz) = 3.8 MHz
INA163 will definitely NOT work in 3V single supply operation since it requires a minimum of 9V (or +/-4.5V) supply voltage - see below.
The only instrumentation amplifier I could suggest that meets you supply voltage (2.7V min) and bandwidth requirement (~1.5MHz in G=31.6) is INA331 (single) or INA2331 (dual). For Vs=3V single supply operation, its common-mode input voltage range is from 350mV to 1.8V while the output voltage range is from 50mV to 2.950V.
Thank you very much for your thoughtful reply !Two things have changed which may make INA 163 a good candidate for our design
1. Supply voltage is between 9V and 12V (and NOT 3V as before), 2. Input signals are differentail and are between 5uV and 10uV (and NOT at 5uV to 10mV as before)
Since input signals are close to noise floor, I have to use a preamplifier to raise those signals to ~1V and then using main amplifier to raise the voltage.All other information stay the same as before.I hope now, you will be able to recommend parts for pre and main amplifier.
I look forward to hear back from you.
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with respect to these materials. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.