Sonsivri
 
*
Welcome, Guest. Please login or register.
Did you miss your activation email?
March 28, 2024, 11:48:44 11:48


Login with username, password and session length


Pages: [1]
Print
Author Topic: looking choice of op-amp for a buffert amp  (Read 3858 times)
0 Members and 1 Guest are viewing this topic.
sphinx
Hero Member
*****
Offline Offline

Posts: 899

Thank You
-Given: 604
-Receive: 260



« on: July 05, 2017, 08:26:10 20:26 »

i am looking for a choice of buffer op-amp for using as input
to an adc converter the voltage will not go above 2 volts
and will not have any fast changes it will be adjusted by
hand so we are not talking even 10Hz i guess.

not so familiar with using them in ad conversion

thanx in advance
Logged

laws of physics are not laws at all, just assumptions and formulas that work as long as we don't figure something new that wrecks the calculations. the infinite onion try to peel that one
optikon
Cracking Team
Hero Member
****
Offline Offline

Posts: 848

Thank You
-Given: 1092
-Receive: 1999


« Reply #1 on: July 05, 2017, 08:42:45 20:42 »

i am looking for a choice of buffer op-amp for using as input
to an adc converter the voltage will not go above 2 volts
and will not have any fast changes it will be adjusted by
hand so we are not talking even 10Hz i guess.

not so familiar with using them in ad conversion

thanx in advance

Depending on the A/D sometimes these have an anti-alias filter in front of them that the opamp has to drive.

Considerations for opamp selection:

1) what A/D resolution & accuracy - surely you do not want your opamp destroying the A/D specifications you care about.
2) DC specs (temp drift, offset / gain errors, linearity)
3) AC specs - Settling time (how fast is A/D conversion / clocked?) - THD, noise etc..
4) Supply rails available, what is you full scale 2V ? Considerations for RR input and RR output, common mode input range of opamp...

The A/D in your system is likely one of most expensive components. Right?
The OPAMP job is a big one to preserve the fidelity of your signal path.

Tell me about your A/D and requirements, I can offer some more advice.
« Last Edit: July 05, 2017, 08:45:29 20:45 by optikon » Logged

I can explain this to you. I can't comprehend it for you.
sphinx
Hero Member
*****
Offline Offline

Posts: 899

Thank You
-Given: 604
-Receive: 260



« Reply #2 on: July 05, 2017, 09:59:05 21:59 »


i choose a MCP3424 since it was supported in ccs and i got it working in proteus. so i can make
a choice between 12-18 bits my choice is not definitive so far. i need to get the pic and the adc
for practical testing so i can see in real life how the circuit behaves.

Logged

laws of physics are not laws at all, just assumptions and formulas that work as long as we don't figure something new that wrecks the calculations. the infinite onion try to peel that one
optikon
Cracking Team
Hero Member
****
Offline Offline

Posts: 848

Thank You
-Given: 1092
-Receive: 1999


« Reply #3 on: July 05, 2017, 10:29:07 22:29 »

i choose a MCP3424 since it was supported in ccs and i got it working in proteus. so i can make
a choice between 12-18 bits my choice is not definitive so far. i need to get the pic and the adc
for practical testing so i can see in real life how the circuit behaves.



Ok, sounds ok, but I do circuit design based on performance requirements and understanding the application needs. Picking ADC because its supported in CCS seems like it is not very important to you so you are just playing around with no requirement. Anything will work for this purpose and you don't need expert help.


Logged

I can explain this to you. I can't comprehend it for you.
mars01
V.I.P
Hero Member
*****
Offline Offline

Posts: 535

Thank You
-Given: 692
-Receive: 1756



« Reply #4 on: July 06, 2017, 08:47:06 08:47 »

@sphinx, if you are looking to use the entire scale of a 18bit ADC and your max amplitude of the signal is 2V, meaning that you are going to use a 2V voltage reference, then you will have about 7uV per step which it's a challenge even to design the PCB itself. You'll have a lot of noise and it will require some hard engineering where the choice of opamp buffer is only the beginning.

But if you don't need performance, and the signal variation is small, then you could do some software oversampling and then dumping the lower digits (where the noise should concentrate) to arrive to the original 18bit result.

And as a medium range opamp buffer, I would choose one that is recommended in the datasheet as a buffer for ADC. One like, OPA2350 http://www.ti.com/lit/ds/symlink/opa2350.pdf but you'll have to deal somehow with the about 150uV offset.

Otherwise, choose one opamp with, at the least, low offset.
Like TLC2652A http://www.ti.com/lit/ds/symlink/tlc2652a.pdf (given the fact that the signal is slow changing, chopper type opamp's could fit in your design).
Logged
sphinx
Hero Member
*****
Offline Offline

Posts: 899

Thank You
-Given: 604
-Receive: 260



« Reply #5 on: July 06, 2017, 09:21:00 09:21 »

thanks for inputs i will try not to use any buffert op-amp and i am not going to use 18 bit but i will try with 16bit
which in reality will be 15bit since adc has a swing -2.048 - 0 - 2.048 but i am only using the positivie half.

if this does not work i will try lower bitrates i could not find a 0-5v swing on an adc that i could get to work
with ccs with 12-16 bit to try software out.

@ optikon. no this is not for any design its just a private project i am working on. i do these since whiplash to
do some things at least to use my head and not stagnate and stop learning. it was over 10-13 years ago i
did any work at all so i do these small projects to do just about to have a bit of fun and to try a bit harder
to learn new things since i like to learn and experience i am learning altium with more and more features
each time i try it i am not in any way any expert on it but its fun to use and learn something new, just to  and
stay sharp, and i appreciate all your input it also teaches me something new and things i didn't think of.

i just picked an adc in ccs since i dont want to write libs and include files to get an adc to work with
with a compiler thats a bit of over my head work. i started with arduino and i then i wanted to try C so
tried ccs c for pics and found a pic fairly cheap and got an adc to play with i will attach for those that
want have a peek at it. the software needs some tuning i am not satisfied with some parts of it, its a
work in progress so far.

many thanx
sphinx
« Last Edit: July 06, 2017, 09:23:51 09:23 by sphinx » Logged

laws of physics are not laws at all, just assumptions and formulas that work as long as we don't figure something new that wrecks the calculations. the infinite onion try to peel that one
Wizpic
Global Moderator
Hero Member
*****
Offline Offline

Posts: 1195

Thank You
-Given: 538
-Receive: 408



« Reply #6 on: July 06, 2017, 02:38:51 14:38 »

I would not use an op-amp buffer this will introduce unwanted noise, I've done a lot of playing with this one and the MCP3426 version, below is the circuit I use where a add a filter to filter out about 10hz the readings become very stable and actuate. ignore the the op-amp part as they boost the voltage from a shunt which can measure +/- signals, could be used for voltage converter just change the values around and calculate for voltage, I also used this in a wireless meter which can measure voltage +/100v. Using a 2.048 vref the output voltage reads half of vref then either adds or subtracts the voltage from vref, if you just wanted to measure 2.00v then just try the filter the. Just use the 2 resistors and capictors I used a 1uf and 10uF ceramic caps.
Logged

When you think, "I can't do anymore. I need a break," that is the time to challenge yourself to keep going another five minutes. Those who persevere for even an extra five minutes will win in life..
bigtoy
Active Member
***
Offline Offline

Posts: 238

Thank You
-Given: 320
-Receive: 297


« Reply #7 on: July 08, 2017, 08:34:52 20:34 »

We do a lot of op-amp circuits driving ADCs at my work. A favorite part is the OPA2333 from TI.

One thing to watch out for, with any op-amp, is to answer the question, "does the output of the op-amp need to go down to it's bottom power supply rail?"  For example, are you powering the op-amp from 0V & 3.3V and you need the op-amp to be able to output a voltage down to 0V. For single-supply circuits this is a common consideration. If you need the op-amp to be accurate down to within a few mV of it's bottom supply rail (0V in my example) then that limits your choice of op-amps. Linear Technology is a good place to look for op-amps that can do that.

The OPA2333 will accurately output a voltage down to within 50mV (less actually) of its lower rail.
Logged
metal
Global Moderator
Hero Member
*****
Offline Offline

Posts: 2420

Thank You
-Given: 862
-Receive: 678


Top Topic Starter


« Reply #8 on: July 08, 2017, 09:58:12 21:58 »

is this caleld rai-to-rail, or this feature has another name?
Logged
h0nk
Senior Member
****
Offline Offline

Posts: 256

Thank You
-Given: 208
-Receive: 230



« Reply #9 on: July 08, 2017, 10:25:35 22:25 »


It is worth to note, that most of the specification is valid for signals around zero
and bipolar supply and can extremely degrade if the rail-to-rail feature is abused.
For maximum (DC-)precision at near zero levels i would always recommend a bipolar supply.


Best Regards
Logged
Pages: [1]
Print
Jump to:  


DISCLAIMER
WE DONT HOST ANY ILLEGAL FILES ON THE SERVER
USE CONTACT US TO REPORT ILLEGAL FILES
ADMINISTRATORS CANNOT BE HELD RESPONSIBLE FOR USERS POSTS AND LINKS

... Copyright © 2003-2999 Sonsivri.to ...
Powered by SMF 1.1.18 | SMF © 2006-2009, Simple Machines LLC | HarzeM Dilber MC