Learn how easy it is to sync an existing GitHub or Google Code repo to a SourceForge project! See Demo


DSP feedback and approach

  • Hi,

    I am having some funny behaviour with the feedback and the approach.
    The electronics works nicely with a gain of x10 in the Z-signal, typical CP,CI are 0.07/0.05, with a piezo sensitivity of 3nm/V.
    But two strange things:
    - To retract the tip, I change the CI value to -0.05, but sometimes (as seen in the Pan-view plugin) the tip moves forward instead of retracting. Is there another way of retracting the tip? Any particular value that should work always?
    - To approach I am using the python script. If I use it right away after loading the DSP program (launching gxsm2 and the sr_spm_control script), it works nicely (I am using the omicron MSCU with the Coolrunner on the DSP). But if I change some settings on gxsm2 (such as the Z-gain) the approach does not work and it gets stuck in stage 2 (AppMd 2, Cp=+0.06,Ci=+0.06, Cire=-0.2991)... with a lot of noise in the current (which seems to be induced from the Z-signal). If I reload everything it works again.

    My DSP code version is SPM version 1057, SoftID 1001, magic EE01.

    BTW, we are planning for a DAC board for the MK2 DSP, without sigma-delta converters. How is the porting of the DSP code to the MK2 going?


    • If your are working in STM mode just set the setpoint to the lowest possible value, eg. 0.0001nA. Keep the parameters of the feedback loop constant. As you have usually some noise on your current signal the feedback loop will retract the tip. By changing the setpoint you can actually get a good value for the noise limit of the whole system. By the way: The threshold when the tip retracts should be the same for both polarities of the bias voltage. If not, you have to check the offset of you SRanger board. You can use the python scrip sr_spm_control.py for this purpose.
      Thorsten :-)

    • Percy Zahl
      Percy Zahl

      >> forward instead of retracting.

      Never seen this.

      >>AW script:
      One issue:

      Never try to do "interfering" actions. I.e. change parameters (CP/CI/Bias/etc...) via GXSM whil runnign the python auto app. -- Always stop the approach, change, then continue. If you change for example CI, just at the time the python script manipuilates it -- things go wrong....

      PS: MK2 -- I am stuck with the USB high speed bulk transfer -- always crashes my whole system, not nice to work on :-( I can communicate in HPI mode just fine. Once the module works, I expect quickly having the DSP code ported and running in a 100% SR compatibility mode.