| Author |
Message |
maalma
Joined: 01 Mar 2007 Posts: 33
|
30 Jun 2009 11:16 passive average circuit |
|
|
|
|
Hello all,
I have 3 input voltages VA, VB and VCM from these voltages I need a circuit that can give me in the output these values VCM+(VA+VB)/2 and VCM-(VA+VB)/2
How can I do this??
Thanks for your help
Last edited by maalma on 01 Jul 2009 13:51; edited 1 time in total |
|
| Back to top |
|
 |
kwkam
Joined: 25 Feb 2002 Posts: 299 Helped: 15 Location: Somewhere on earth
|
30 Jun 2009 16:06 passive averaging circuit design |
|
|
|
|
| you can use opamp approach or V to I method.
|
|
| Back to top |
|
 |
maalma
Joined: 01 Mar 2007 Posts: 33
|
30 Jun 2009 16:52 passive averager circuit |
|
|
|
|
alright, can you give me more details please
thanks for your reply
|
|
| Back to top |
|
 |
AdrianN
Joined: 28 Jun 2009 Posts: 20 Helped: 10 Location: California
|
|
| Back to top |
|
 |
maalma
Joined: 01 Mar 2007 Posts: 33
|
02 Jul 2009 9:29 Re: A passive circuit that can average 2 voltages |
|
|
|
|
Thank you Adrian for your help
i'll try your solution and i'll keep you informed if it works in my application
actually, resistors are very "noisy" and i'm using this in an ADC so the noise must be <1/2LSB=12mV.
i wonder if i can replace the resistors by capacitors??
thanks again
|
|
| Back to top |
|
 |
Google AdSense

|
02 Jul 2009 9:29 Ads |
|
|
|
|
|
|
| Back to top |
|
 |
FvM
Joined: 22 Jan 2008 Posts: 5177 Helped: 768 Location: Bochum, Germany
|
02 Jul 2009 15:10 A passive circuit that can average 2 voltages |
|
|
|
|
| Curiously, the question title requests a passive circuit, however all suggestion are about active solutions. That's plausible in so far, as a passive subtractor isn't feasible. Also SC is a possible solution, but not regarded as passive usually.
|
|
| Back to top |
|
 |
AdrianN
Joined: 28 Jun 2009 Posts: 20 Helped: 10 Location: California
|
03 Jul 2009 18:42 Re: A passive circuit that can average 2 voltages |
|
|
|
|
I would not be worried about the resistor noise. Johnson noise on a 1kohm resistor is about 4nV/sqrt(Hz) at 25degC. Even if you have a bandwidth of 100MHz, the resistor noise would be 40uV. There are 10 resistors in this circuit and with a gain of 1 (or around) the noise generated by these resistors is roughly 40uV*sqrt(10) = 126uV, 40dB below your 12mV requirement.
Moreover, you can choose low noise resistors like metal-film or thin-film resistors, and choose a lower value resistor (hundreds of ohms) if the power consumption is not an issue.
I would be more worried about the flicker noise of the Op Amps you will use, but because the gain is low, that might not be an issue as well.
Therefore, use low noise resistors, low noise Op Amps, have a continuous plane ground to reduce rf interference, use surface mount parts, short traces, good decoupling, and reduce the bandwidth to the minimum required, and you should easily achieve below 12mV noise floor.
|
|
| Back to top |
|
 |
maalma
Joined: 01 Mar 2007 Posts: 33
|
04 Jul 2009 9:33 Re: A passive circuit that can average 2 voltages |
|
|
|
|
thank you Adrian for these explanations
can you please tell me more about a low noise opamp (i'm really a new commer in the analog design world)??
have a nice day
|
|
| Back to top |
|
 |
AdrianN
Joined: 28 Jun 2009 Posts: 20 Helped: 10 Location: California
|
|
| Back to top |
|
 |
maalma
Joined: 01 Mar 2007 Posts: 33
|
05 Jul 2009 9:24 Re: A passive circuit that can average 2 voltages |
|
|
|
|
Hello Adrian,
Thanks for the link, but i'm not talking about devices, i'm talking about schematics to simulate with cadence.
anyway thanks for your help and welcom to EDA board
|
|
| Back to top |
|
 |