| Author |
Message |
birdqiang
Joined: 28 Jul 2004
Posts: 19
|
 01 Sep 2004 3:36 about offset of amp |
|
|
|
|
in what condition the offset voltage is taken into acount?
and how to lower it ?
|
|
| Back to top |
|
 |
Humungus
Joined: 10 Jul 2001
Posts: 418
Helped: 26
|
| 01 Sep 2004 7:41 Re: about offset of amp |
|
|
|
|
Offset is important in all applications where the DC level plays a role.
Also remember that you have two offset components: systematic and random.
Systematic offset arises for imperfections in your design. Random offset is given by the random mismatch appearing among devices "supposed" to be equal.
Reducing systematic offset depends upon your architecture. You can correct it by properly biasing your transistors and by appropriate setting of passive components.
Random offset mostly depends on the size of your devices and how they are laid out in your design. Also, there is a part of random offset depending on your bias, as active device mismatch is bias dependent.
Another way to reduce offset is the introduction of switched techniques, such as chopper, ping-pong and auto-zeroing.
|
|
| Back to top |
|
|
neelandan
Joined: 26 Mar 2004
Posts: 28
Helped: 1
Location: Terra Firma
|
| 01 Sep 2004 7:51 about offset of amp |
|
|
|
|
Offset voltage is generally specified as the equivalent input voltage. That is to say, the generated output voltage divided by the gain.
The offset is due to two reasons: one, the difference between the input voltages of the two active input devices, commonly called the voltage offset, and,
two, the offset due to the difference in current drawn by the two inputs.
The voltage offset has to be nulled out: It is compensated by adding an equal and opposite voltage.
The offset due to the input currents is minimised by ensuring that both inputs of the opamp see the same resistance.
Depending on the application circuit, the problem of ensuring minimal offset can get quite complicated.
In most cases, however, the op-amp manufacturer will have a recommended scheme for nulling his op-amp.
The resistance from each of the inputs have to be (roughly) equal.
|
|
| Back to top |
|
|
watersky
Joined: 20 Aug 2004
Posts: 70
Helped: 2
|
| 02 Sep 2004 1:54 about offset of amp |
|
|
|
|
| using autozeroing techniques cancels offset,you can find is from page 464 in ALLen book.
|
|
| Back to top |
|
|
birdqiang
Joined: 28 Jul 2004
Posts: 19
|
| 02 Sep 2004 5:22 thanks to Humungus |
|
|
|
|
| Humungus wrote: |
Offset is important in all applications where the DC level plays a role.
Also remember that you have two offset components: systematic and random.
Systematic offset arises for imperfections in your design. Random offset is given by the random mismatch appearing among devices "supposed" to be equal.
Reducing systematic offset depends upon your architecture. You can correct it by properly biasing your transistors and by appropriate setting of passive components.
Random offset mostly depends on the size of your devices and how they are laid out in your design. Also, there is a part of random offset depending on your bias, as active device mismatch is bias dependent.
Another way to reduce offset is the introduction of switched techniques, such as chopper, ping-pong and auto-zeroing. |
thanks a lot! I know zuto-zeroing techniques, but don't what is the chopper and ping-pong?
|
|
| Back to top |
|
|
Humungus
Joined: 10 Jul 2001
Posts: 418
Helped: 26
|
| 02 Sep 2004 8:25 Re: about offset of amp |
|
|
|
|
Classical chopper means that you "chop" you input signal by multipying it with another signal which changes its value periodically from "1" to "-1" and then you do the same with your output signal. As offset internally to your system doesn't change its sign, the output signal (after the output chopper) is the signal you should have ideally, plus the system ofset multiplied by a an alternating series of +1 and -1. So, by filtering out the varying offset signal, you can recover the output signal with no offset. The multiplying operation by +1 and -1 is performed by reversing the polarity of the input and output signals by means of switches.
Another technique also called chopper stabilizations consists in having two amplifiers, each one with an anciliary input to cancell offset. One amplifier, called the main amplifier is always connected to the signal path. The other amplifier, called nulling, makes an auto-zeroing step to cancell out its own offset, it store it into a capacitos and feeds this offset to its own anciliary nulling input. Then it looks for the signal at the input of the main amplifier and apply a correcting voltage at the nulling input of the main amplifier, storing also this voltage into another capacitor. The auto-nulling and main-nulling steps are repeated periodically. You can look for information about this method by searching "zero offset amplifiers". This is the method used by commercial amplifiers called, zero-offset or zero-drift by burr-brown, analog devices, etc. As you always cacell the offset at any moment, the offset drift with temperature (maybe the most anoying fact of offset) is removed. Properly implement this method requires a lot of analysis, due to there are many tradoffs between gains, bandwidths, chopping frequency, etc.
Ping-pong consists in having two identical amplifiers which, each one at its turn, connects to the signal path, while the other is in the auto-zero stage. Then the amplifier which was at AZ, connects to the signal path, while the first one makes its AZ and so on.
|
|
| Back to top |
|
|
yyliang
Joined: 26 Aug 2004
Posts: 43
|
| 02 Sep 2004 13:59 Re: about offset of amp |
|
|
|
|
The offset of amplifiers is mainly caused in the process of manufacture,to
BJT TECHNOLOFY,the value of offset is about 1-2mV,to CMOS technology,the value is about 20mv,the offset can be canceled ,you can see in page464-466 of CMOS Analog Circuit Design By P.E.Allen and D.R.Holberg.
|
|
| Back to top |
|
|
ablue
Joined: 22 Jan 2003
Posts: 40
Helped: 3
|
| 05 Sep 2004 15:21 about offset of amp |
|
|
|
|
we can eliminate system offset by properly biasing
If we dont have autozeroing in your system , we just can minimize the process offset by setting the size of XTORs large and properly layout them, because we cant predict process offset
|
|
| Back to top |
|
|
Bluechem
Joined: 05 Sep 2004
Posts: 71
Helped: 1
|
| 06 Sep 2004 3:00 about offset of amp |
|
|
|
|
| different op-amps have different offset voltage. if you need very low offset you should use a precision op-amp. also keep in mind that some times offset voltage come from input voltage not op-amp itself(for example a drift in voltage of a sensor due to temprature changing)
|
|
| Back to top |
|
|
frankon_o2
Joined: 07 Sep 2004
Posts: 8
|
| 25 Sep 2004 5:23 about offset of amp |
|
|
|
|
| to reduce offset of opamp, we can improve the matching of input mosfet pair by draw them in cross coupled style.
|
|
| Back to top |
|
|
mandrei
Joined: 10 Jan 2004
Posts: 67
Helped: 2
|
| 02 Oct 2004 7:11 Re: about offset of amp |
|
|
|
|
Just some notes about offset:
Supose in data book you read : offset for the OpAmp is 30mV
In simulation this mean that you must consider a voltage source between -30mV...+30mV at your + input of the OpAmp and observe the DC effect at output (Observe offset can be also zero in simulation-ideal case!!)
Than consider the same source at the - (minus) input of your OpAmp and observe the effect at output.
Usualy offset is considered only in worst case: that is +30mV and -30mV
|
|
| Back to top |
|
|
