Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Low cost amplifier for 12bit ADC SAR?

Status
Not open for further replies.

Coper

Member level 4
Joined
Mar 16, 2017
Messages
69
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
773
I have 1MSPS 12bit ADC SAR for measure voltage between 50mV - 2500mV. Because signal source have impedance about 100k ohm I have to use an amplifier as buffer.
I use 2 MHz Zero-Drift Op Amps MCP6V71U in SC70-5 after the technical site is all OK , but price this amplifier is over 1$ and that's expensive.
Do you advise a cheaper amplifier suitable for 12bit ADC?
 

Hi,

"12 bit ADC" here is no useful specification.

You need to specifiy
* the expected error (your personal decision)
* the sampling timing (ADC datasheet)
* the load behaviour from ADC to the OPAMP (ADC datasheet).

Klaus

Added:
You surely know that one LSB may be about 0.5mV.
Thus the OPAMP zero drift of +/-15nV/°C is ovekill. --> This equals to about 33000°C per LSB!!

What temperature range do you expect?
Divide this value by the expected error to decide the V_OS drift specification of the OPAMP.

Klaus
 

OK
I use internal ADC in STM32F373 with external reference (TL431).
I use sampling gite 960ns or more.
Already from the ADC used and the reference is clear, there is no extra accurate measurement,my idea is sum of all error inserted in amplifier must be lower than 2LSB it is 1200uV (temperature drigt 0-50C + input offset error ) of course I will welcome if I enter into 1LSB
 

Hi,

my idea is sum of all error inserted in amplifier must be lower than 2LSB it is 1200uV
Then it makes no sense as TL431 has
* +/-2% accuracy error, which equals to 163 LSB
* and with 6mV temperature drift error, which equals to 10 LSB

The ADC error will also be in the range of +/-3LSB.

***
The ADC switches a 8pF load (hold capacitor) to the OPAMP´s output. This causes ringing with slow (low power) OPAMPs.
This ringing amplitude decreases with time. You need to ensure that ringing is below your expexcted LSB error befor the capacitor is dsiconnected from the OPAMP and conversion starts.
Thus a 100R, 470pF lowpass from OPAMP output to ADC input should be a good way to prevent ringing.

What´s the input signal maximum frequency you want to convert with the ADC?

Klaus
 

In a way you are right, but
my measurement is Ratiometric and this remove error of reference ,
Yes no not remove error of ADC , but it's no reason to add more big error on buffer amplifier.
I'm not talking about the absolute precision +2LSB, but only an error arising on buffer amplifier.
 

Hi,

my measurement is Ratiometric and this remove error of reference ,
This makes sense.

***

"absolute precision" --> is called "accuracy"
"repeatability" --> is called "precision"

Klaus
 

Accuracy and precision

Prec.png
 

There are a million opamps out there and there's a reason there are a million opamps (something I've slowly come to appreciate) and it's because every specification matters for a given application.

So to help we'd need more details like signal bandwidth (not necessarily related to sample rate), voltage range, power requirements, application temperature range, input impedance?


There may be other system level options too
-Can you periodically switch out/short out the measured signal to record and null offset? If yes then amplifier offset isn't important at all
-Can you record and store (flash) initial offset during manufacturing? If yes then offset drift is all that matters
-How wide is your operating temperature range? If its narrow then 1mV of offset is quite easy for non chopping amplifiers
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top