• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 48 subscribers
  • Views 903 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Strange noise model for LM6172

Philippe MOREL
Prodigy 10 points
Other Parts Discussed in Thread: LM7372, LM6172, TINA-TI, LM6171

hello,

I'm trying to compare noise performance of 3 similar Single ended to differential circuits.

The basic circuit is always the same, just the amplifier itself is changed See SE2DIFF_LM6172_LM7372_LT1364.EMF in Word document

case 1 uses LM7372

case 2 uses LM6172

case 3 uses LT1364 (from LINEAR)

See comparative noise performance.emf in Word Document

Case 1 noise analysis from TINA-TI (Version 9.300.244 SF-T) gives quite expected results. the 1/f noise increase of the Amplifier is clearly observable

Case 2 gives strange results : No 1/f noise and a very strange noise increase above 50kHz !?

Case 3 is also very strange resulting in flat & very high noise.

If I run a similar noise analysis with LT-SPICE, the results look much more like expected. see SE2DIFF_LT1364.pdf

Questions :

Why is case 2 (LM6172) so strange ?

Why is case 3 also so strange and so different from LT-SPICE ? I did import the model from LINEAR into TINA-TI

Thanks for your anwer

Noise analysis issues.docx

3125.SE2DIFF_LT1364.pdf

  • 0
    TI__Mastermind 22825 points

    Hi Philippe,

    Sorry about the late response, as your E2E post was not in a place I normally monitor and respond regularly.

    I believe you've found that the LM6172 modeled input noise (voltage and current both) are to blame for the unexpected results / increase at relatively higher frequencies and results being devoid of the expected higher 1/f component at the lower frequencies.

    I'll take up the question with our internal simulation group to see if we can correct the LM6171 (and LM6172) models to get rid of this error. I am not 100% certain that this can be done as the device is rather old and the original designer / macro-model developer and have since moved on to other things. Either way, I'll post my findings about this in this post for future reference.

    Sorry about this error in the model and I will do my best to see if we can correct the issue.

    Here are some supporting files:

    /cfs-file/__key/communityserver-discussions-components-files/14/8030.LM6172-LM7372-Noise-Compared-3_5F00_16_5F00_15.zip

    Regards,

    Hooman

  • 0 in reply to Hooman Hashemi
    TI__Mastermind 22825 points
    Hi Philippe,
    I ran the LT1364 in LT Spice and saw that the output noise was reasonable (about 27nV/RtHz flatband level and shows the 1/f noise noise increase at around 1kHz).

    I tried to import the LT1364 model into TINA-TI but I could not because there were lines in the model file which TINA-TI did not recognize (e.g., 2nd line: "A1 2 1 0 0 0 0 0 0 OTA g=0 in=1p ink=200"). So, if you need further help with the behavior of TINA-TI with LT1364, please upload the *.tsc file you have for it and I'll take a look.

    Regards,
    Hooman
  • 0 in reply to Hooman Hashemi
    TI__Mastermind 22825 points
    Hi Philippe,
    I've heard back from the TI modeling folks that they will update / correct the LM6172 model and have placed it in their queue for new development. Unfortunately, I don't have a proposed completion date to share.

    I will post any updates here for future reference.

    Regards,
    Hooman