Noise in a simple circuit

Discussion in 'General Electronics Chat' started by Alex009, Jul 30, 2014.

  1. Alex009

    Thread Starter Member

    Jul 2, 2014
    30
    0
    Hey guys it's me again,

    I got a question conerning one of my circuits. Its a pretty simple one.
    I m using a LT1227CN8 (OPAMP). I m using it as an inverting op. So there is an 1 MOhm resistance on the - input and 1 MOhm ressistance between the output (i know thats a lot). Between the positive input and ground i got a 499kOhm ressistance. On the output i got a 330pF capacitor.

    So if do some math i get to a noise of (about) 80µV under the following conditions:
    Temperature: 300K
    Max. Frequenzy 150kHz
    I also assume that the typical number given in the datasheet is constant over the the whole bandwith to 150kHz (min Frequenzy 1kHz). Which is obviously not the case since it gets lower at higher frequenzies. And higher at lower frequenzies.

    My questions would be:
    a) Are these assumptions viable in reality?
    b) It doesnt seem to me like the cirucit has an -1 Gain (after measuring it in reality - and I dont get why)
    c) The noise seems a lot higher than the calculated one. It also gets a lot higher as soon as the OPA starts working (gets his supply voltage).

    Cheers, BR
    Alex


    /edit: the supply voltage I m using is +/- 12 generated out of the following DC/DC converter: TMR1-2422 (i used some capacitors and inductive elements to flatten the supply voltage - if that makes any sense to you)
     
  2. Alex009

    Thread Starter Member

    Jul 2, 2014
    30
    0
    No idea? Not even according one of these questions? :( Or did i just express myself poorly?
     
  3. daviddeakin

    Active Member

    Aug 6, 2009
    207
    27
    The data sheet: http://cds.linear.com/docs/en/datasheet/1227fb.pdf
    says the inverting input has a current noise density of 32pA/rtHz :eek:
    This flows in the Thevenin source resistance of 500k, making 16uV/rtHz, which when integrated over 150kHz makes 6.2mV noise! No wonder your circuit is noisier than you expected! :p
     
  4. crutschow

    Expert

    Mar 14, 2008
    12,995
    3,229
    The op amp itself has noise.

    Also note that the LT1227 is a wide-bandwidth, current-feedback type op amp and won't work properly with such high impedances. Current-feedback op amps also tend to have higher noise. I suggest you read the spec sheet of the devices you use. It'll save a lot of questions. :rolleyes:

    For your configuration you need a low-noise, low input bias-current, voltage-feedback op amp.
     
  5. Alex009

    Thread Starter Member

    Jul 2, 2014
    30
    0
    Boody hell, your right. I just read 3.2 I guess ..now i feel kinda stupid :rolleyes: Thats why math and reality dont match at all..cheers for pointing it out. Well..for my defense..after reading hundreds of datasheets i prob got tired.
     
  6. AlphaDesign888

    New Member

    Jul 27, 2014
    189
    10
    Bloody hell? You sound Australian. Here's another one from Australia 'bloody poofters'
    Got a joke for you ...

    What did the Japanese poofter say?
     
  7. Alex009

    Thread Starter Member

    Jul 2, 2014
    30
    0
    Well actually i think i got this "Bloody hell"-thing from the British guys since i lived there for a while. But whatever.. i got no idea :)
     
  8. AlphaDesign888

    New Member

    Jul 27, 2014
    189
    10
    The dirty Japanese poofter say 'arsesore'
     
  9. AlphaDesign888

    New Member

    Jul 27, 2014
    189
    10
    Stupid me. Of course you're British you say 'an' not 'a'

    An LED
    A LED

    Which one sounds the most appropriate to all you guys?
     
  10. Alex009

    Thread Starter Member

    Jul 2, 2014
    30
    0
    According to my passport I m german, but lived in England and Canada :p .. so yea you could be right. Back to topic.. could the LT1122 work?
    http://cds.linear.com/docs/en/datasheet/1122fb.pdf
    If memory serves correctly i already tested this one but i think i stoped using it since it has a strange voltage gain over frequenzy curve (used in the above circuit)
     
    Last edited: Jul 31, 2014
  11. AlphaDesign888

    New Member

    Jul 27, 2014
    189
    10
    I cannot see why not. Perhaps an expert can shed some more light on the matter.
    My level of knowledge is very 'generic'. What we need is an expert with 'specific' knowledge.

    Canada? Gawd I'd be doing 3 to 5 in prison for drink driving. If we had these consequnces in Australia then I would not have gotten behind the wheel! I lost my license for 12-months
     
  12. crutschow

    Expert

    Mar 14, 2008
    12,995
    3,229
    Note that some op amps don't like a capacitive load (such as the 330pF you mentioned in your setup). Why did you add that as it serves no useful purpose other then simulating a high capacitance load? It doesn't significantly reduce output noise.
     
    Alex009 likes this.
  13. Alex009

    Thread Starter Member

    Jul 2, 2014
    30
    0
    Oh well i should explain that: I m trying to get a certain amount of charge at the output (on the capacitor) . Thats the actual goal. And it needs to be inverted, thats why i have to use an opamp. And on top of that i need an output impedance which is as high as possible and low noise on the output signal - which sounds like thats gonna be hard to achieve. (Maybe its just hard for me)

    These requirements are all related to a following device.
    If you got any other idea achieving these requirements you are very welcome to give me some inisght in your idea.
     
    Last edited: Jul 31, 2014
  14. crutschow

    Expert

    Mar 14, 2008
    12,995
    3,229
    Perhaps if you explained the application, we could offer better suggestions.
     
  15. Alex009

    Thread Starter Member

    Jul 2, 2014
    30
    0
    [​IMG]Could the following OP Amp work with such a high capacity load: http://www.analog.com/static/imported-files/data_sheets/AD797.pdf since this one meets all the other aspects. Sadly I m not having any experience with high capacity loads (well the good thing about this is: i can learn a lot new stuff)
    ? If used in combination with a compensation path?

    Well i know we drifted apart from the original topic - but I dont wanna start another thread for it.

    // edit

    added the measurement results.. granted i dont have to many points but its enough to give you an overview (at least i hope so)

    http://www.pic-upload.de/view-24164818/LT1028.png.html
     
    Last edited: Aug 7, 2014
Loading...