运放和电压源的稳定性以及精度

只看该作者 · 2011-3-22 02:31

最近想用电阻分压后挂跟随器构成电压基准源。
现在忽略电阻精度。
对于运放而言，影响电压源精度最重要的是什么？input offset ? out put offset?或者其他
影响稳定性的，我能想到psrr，那么cmrr的影响有多大？
期待maychang老爷子回答：）

只看该作者 · 2011-3-22 10:04

电压基准源. 这个貌似和运放没啥关系。

用电阻肯定不精确的。电源波动一下，你分压出来的电压值就动了。
电压基准源说白了就是带隙电压源 1.205V，这些电路的PSRR通常>80dB 甚至在100dB以上，而且对于温度变化的影响也微乎其微，通常在1ppm/ ℃.
要是用分立元件来组建的话，这个得花些功夫调研一下。

1万 · 2011-3-22 10:26

本帖最后由 zjp8683463 于 2011-3-22 10:28 编辑

如果基准用与ADC等不要用分压，要用基准芯片

分压主要的问题在于稳定性差。电阻和电源的漂移是最大的问题，运放的漂移相比来说很小。

如果能接受100ppm的漂移，就可以用分压。

只看该作者 · 2011-3-22 14:25

No, I think you have misunderstood what I mean...
We could ignor all the influence from the resistor, just discuss the results which could be mainly determined by the parameters of OPA.
In other words, if we use a voltage source IC to generate the reference voltage, then add a OPA to enlarge the output current, the OPA parameters have if infuluence on the output, so, which parameter(s) would be more important? What is the infuluence by then?

3万 · 2011-3-22 15:06

4楼：
我估计你是用电阻从电压基准源输出处分压，得到电压值不同的一个基准，再用运算放大器扩流。
暂不考虑电阻的精度和温漂，那么这种情况下，影响输出精度的，主要是运放的失调电压，影响输出稳定性的，主要是运放的失调电压温漂。

1万 · 2011-3-22 16:04

对于运放而言，影响电压源精度最重要的是什么？input offset ? out put offset?或者其他影响稳定性的，我能想到psrr，那么cmrr的影响有多大？

input offset和output offset是统一的，失调（电压）可以折合到输入端也可以折合到输出端。
运放的精度应该受到offset影响，但到底是电流offset还是电压offset影响更主要，往往与信号源有密切关系。至于最重要的是谁恐怕很不好说，得看应用场合吧。

只看该作者 · 2011-3-22 16:19

5# maychang

Thanks a lot!
You are right, do you remember the AD620 single voltage source application on bridge sensor? My problem is very similar to it.

Would you please show me in detail on the offset voltage? Input offset or output offset?
Another problem was how serious the influence related to the CMRR? On the PCB, we have strong noise from the BLDC and so on, they would lead a common mode noise on all the circuit, in the follower, I think its CMRR would be lower, maybe it will lead to a big noise for the final output?
Cause I am in the trip to our head quarter, in the working time there was noChinese input methode, I am sorry to make you read in English.

1万 · 2011-3-22 16:35

上班不能打汉字，意味着什么？

只看该作者 · 2011-3-22 17:07

外国人啊。我英文不好。学英语真难啊，学了十来年感觉还在起步阶段。

只看该作者 · 2011-3-22 18:35

8# iC921
Thanks a lot, I have seen some datasheet which give both the input and output offset specification, so I have this question.
In this case, I think the signal source could be described as DC, low source impedence, maybe the offset voltage would contribute more to the final output?

PS: I am in Europe now :)

只看该作者 · 2011-3-22 18:35

8# iC921
Thanks a lot, I have seen some datasheet which give both the input and output offset specification, so I have this question.
In this case, I think the signal source could be described as DC, low source impedence, maybe the offset voltage would contribute more to the final output?

PS: I am in Europe now :)

只看该作者 · 2011-3-23 00:47

本帖最后由 bbyeah 于 2011-3-23 00:49 编辑

The first problem is if you choose resistive divider, there is no PSRR compare to real bandgap reference.
Assuming these two resistors has the same temperature coefficient and long term drift, then the only time and temperature depedence is from OPA.
Since OPA has almost infinity PSRR compare to resistive divider in low frequency, this is not so important, but you should consider the input bisa current, offset voltage/current, sure the time/temp drift is the main parameter which will effect your measurement. The offset can be measured and compensated, but drift can not be.

Voffset,out=A*Voffset,in
for most CMOS OPA, with nominal input offset voltage, the output voltage is saturated.