Steve, 1 msec is not per scan, it's per step. One scan line of 1000 steps takes 1 second. The period of the sawtooth function is 1 second. 1000 lines takes 1000 seconds or about 16 min.So ... you want to use the output of a cheap DAC to generated a sawtooth waveform to drive a piezo transducer to create linear movement to move something under a microscope. You want this to be accurate, position repeatable and scan moving at 1mSec per scan. There's a lot going on there.
In the post that I cited, a noise of 20 uV was supposedly achieved from this cheap DAC (much less than 5V/4096/2 cited here). 20 uV fits my application with an order of mag to spare. 5/4096/2 also might work. If the DAC proves insufficient despite measures I take, I will get a better one.
I am going on a tangent (sorry), but re: nanopositioners. 16-bit and 24-bit DAC used in commercial nanopositioners drive actuators in flexure stages that mechanically amplify the PZT displacements by 10-fold. Users of these want to have 50-100 micron range with sub 1 nm resolution and use these to mount slides and 96-well plates (1x3" or 4x3"). I only want 1 micron range with 1 nm resolution, and I am not using flexures, i.e. the task is simpler than what's done in commercial stages. The resonant frequency of my piezo is 350 kHz and I am operating at 1kHz max (and as we spoke before, 1kHz is a nominal value). I am not worried about lorries etc, b.c. I've used nanostages next door to unbalanced floor centrifuges spinning 1L bottles, and have circumvented vibrations etc.
Statistics: Posted by runcyclexcski — Wed Dec 31, 2025 5:18 pm