DSP new features requests (MK2-A810)

  • Wolfgang Theis

    Wolfgang Theis - 2010-01-12

    I intend to control an MFM (tuning fork AFM based) and an SEM concurrently on
    a single GXSM/MK2-A810 platform. The SEM beam will be used for scanning SEM
    images, Auger point spectra (possibly Auger maps), point RHEED diffraction,
    simple geometry e-beam lithography (lines, boxes).

    For MFM a lift off I would like to aim at a dual-scan mode run on the DSP.
    I.e., record z(x) with feedback for a line and retrace line with z(x)+offset
    with feedback off. A choice for the retraced line on the back scan or as a
    second forward scan would be useful. Percy is looking into the feasibility of
    this (using the DSP external memory).

    For concurrent or alternating operation of MFM (scanning with feedback option)
    and SEM (scanning without feedback option) adding a second DSP scanning task
    and an instrument 1 or 2 parameter for the move and VP tasks would make two
    instrument operation much more versatile and transparent. (Maybe having
    separate move1, move2 tasks instead of move(inst=1,2; ...) would be more
    appropriate. Same regarding VP).

    Concurrent scanning would not be required in any experiments. However,
    STM/AFM/MFM approached with feedback on while SEM scanning would be useful.
    There might be an issue with the input channel averaging which would have to
    be matched to the instrument the channel is assigned to. Options seem to be
    either a duplication, updating V(ADCnumber, assumed instrument), or updating
    V(ADCnumber) according to assigned instrument. I realize all of this, while in
    principle fairly straight forward, would imply major changes of the DSP/HwI
    interface and quite a bit of coding.

  • Wolfgang Theis

    Wolfgang Theis - 2010-01-13

    Something occurred to me:

    Can I keep the MFM feedback running and at the same time change the x and y
    assignment from MFM to SEM DACs and run an SEM scan? I guess that should be
    possible. (MFM/z and MFM/phase shift could always remain on their DAC/ADC). If
    it is possible, I could switch back and forth between instruments on the HwI
    side without any need to change the DSP code. MFM approach would probably also
    work while SEM scanning.

    Is it possible to set VP parameters such that the VP does not change z or
    feedback on/off? I'd want to make sure the tip doesn't crash as a side effect
    when using the SEM.


  • Thorsten Wagner

    Thorsten Wagner - 2010-01-13

    VP is very useful. And yes, you can define feedback on/off as part of a
    vector. You just have to define the vectors you need.

    About your DAC assignment: With the SR-MK2&A810 we actually use for the first
    time different DAC for offset and scanning. In former times the offset was
    (mostly) added to the scan signal on DSP level. That was a cheap solution,
    because you did not need an external amplifier/adder. Another reason was the
    qualitiy of the DAC on the SR-STD. There was no means to have a scanning
    signal with a low gain giving a high resolution and then add to that the
    offset with a high gain as the DAC was unstable on a sub-Hz time scale. So you
    may just use the DACs for the scanning signal for your MFM and the one for the
    offset for the SEM.

    Thorsten :-)

  • Wolfgang Theis

    Wolfgang Theis - 2010-01-13

    Is there an option to disable the offset to free the offset DACs for other use
    and have the SR-MK2&A810 output the absolute x-position through a single DAC?
    What is the current policy regarding x/xoffset? Do move and center scan change
    xoffset to achieve x as close to zero as possible? Are scans always centered
    on x = 0, do VPs have only x as a parameter, i.e. leave xoffset at previous


  • Thorsten Wagner

    Thorsten Wagner - 2010-01-14

    The DACs for the offset are free/unused if offset adding is turned on. Once I
    have used that to control a ramp generator (with the elder SR-STD). For the
    old card you had been able to set for each channel an individual offset to
    compensate for the offset of the internal amplifiers. You can access this
    offset by a python-script. As the new board is much better you do not need
    that any more.

    Internally you handle scan offset and scan signal separatelly. Just at the end
    of the pipe the will be connected to the DACs. The offset adding is just a
    software switch. If on offset and scan signal will be added and connected to
    the DACs for the scan signal. If off, offset will go to the offset DACs and
    scan signal will go to the scan DACs. There is a separate selector for the
    external gain of the offset. But usually you will use the highest gain on the
    offset signal.

    The offset is set by the main window. There will be a ramp to go smoothly to
    the final position. For the scan you can center it to the point given by the
    offset. If not you will have the middle of the top line at the offset point.
    All rotations will be carried out according to the (0,0) position of the scan.

    Spectroscopy is done at the position of the tip. So you can to spectroscopy
    even during scanning without interupting the scan. So if you want to take a
    spectra at a certain position in a scanned image, use the point marker to move
    the offset. If you want to have the same image scanned again, you will have to
    write down somewhere the previous offset settings and then re-enter them
    again. For now there is no button to take a spectra at certain points. That
    might be something useful and the VP engine is for sure able to do the job.

    Thorsten :-)


Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.

No, thanks