User:Steven J. Koch/Notebook/Kochlab/2009/07/15/Feedback modules

Steve Koch 18:07, 15 July 2009 (EDT): Borrowed a USB DAQ board from Sasha (Thanks!) and now trying to update NI-DAQmx (to version 8.9) so the board can work on the old machine. Since the power modulation worked pretty well yesterday (perhaps not fast enough, but we'll see), I want to implement a variety of things that will allow us to use velocity clamp and other clamps. Things to do:


 * Does power modulation give X, Y signals? If so, we need "create offset array" back in the mix (ugh)
 * What are good integer value setpoints?
 * Larry & Linh are making stuck beads so we can look at DOGs
 * get Find Tether Center working
 * Probably requires working force clamp
 * get force clamp working
 * get velocity clamp working

Laser power modulation

 * Laser turned on, 10.31 amps setpoint (actual 10.17)
 * Set power mod control voltage to zero using MAX
 * Set power mod internal setpoint such that transmittance was about 36%
 * 0V = 148.6 mW
 * 1V = 37.5 mW
 * -1V = 260 mW
 * 0V repeat = 148.6
 * (NOTE: laser still warming up probably, so these numbers not too accurate)

Detector power study
0V control, power is 148.6. Sum signal fluctuating around 1.33.


 * 0V = 148.6 mW = 1.33 V
 * 1V = 37.51 mW = 0.328 V +/- 0.001 V
 * -1V = 259.9 mW = 2.32 V +/0 0.01 V NOTE it appears more unstable than above two instances

OK, detector X and Y zeroed at control voltage of -1


 * -1V, X = -0.006 V +/- .002 (noise) .004 (drift)
 * 0V, X jumped to -0.001
 * 1V, X jumped to +0.001
 * back to -1V, X jumped to -0.003
 * back to 1V, X jumped to 0.001

Well, at first glance, this seems REALLY good to me. I don't see any signs of an offset problem. If the beam were actually moving on the detector, I think we'd see huge changes to these numbers. Better check Y too, though:


 * 1V, Y -0.004 V
 * -1V, Y -0.004 V (but drifty at first about +/- 0.003 V
 * 0V, Y -0.001 V
 * 1V, Y -0.002 V
 * -1V, Y, -0.008V
 * 1V, Y, -0.002 V

Slighty worse than X, but maybe due to zeroing being off (remember, we DO expect X and Y raw detector voltages to depend on Sum signal, since the detector does not do the division on-chip.

Front panel control of piezo and power
Steve Koch 04:24, 16 July 2009 (EDT): Holy Shit! I just spent a couple hours overhauling the feedback software. I put in the laser power modulation control, and everything compiles. Furthermore, I updated the front panel control stuff. So far I have only tested out that. But it works! You can now control the piezo and the laser power via the front panel controls:
 * Laser power: 0 V is the minimum. A jump of 1 V is a lot (about 200 mW). Positive number increase = laser power increase.
 * Piezo: Uses AOD center frequency (35 MHz) as "0V". Jump of 1 MHz is about 1.4 microns.

Feedback stepping
Steve Koch 04:29, 16 July 2009 (EDT): Still on a lucky roll: the stepping from step to step seems to work also. See the feedback program in data file 44 (July 09 Calibration).
 * The thing I don't understand now is how the laser power reverts back to low power after the program. Maybe this is a feature of the software that I don't remember?  Is it time to try Find Tether Center?  Nah, I need a DOG first?

Took a DOG curve in files 45 and 46
45 was at voltage of 1 V. 46 at voltage of 0.2 V.
 * file 45: Extent of curve is about +8000, -7000. Very wiggly even with no bead...not sure why that would be.  Doesn't look too shabby, though.
 * file 46: Much, much noisier. Also shifted.  Extent of curve probably about the same.   Looks shitty though.

For now, going to say that a good setpoint integer is 6000

Find Tether Center
Steve Koch 04:47, 16 July 2009 (EDT): First attempt failed. Stuck in loop for some reason (piezo not moving either)

Steve Koch 05:05, 16 July 2009 (EDT): Piezo sensor is not working. Doesn't appear to be working on either DAQ computer. Not good. Working fine on regular oscilloscope. Grounding issue? Working fine on USB DAQ now...cable? Now have it all hooked up with a new cable and seems to be working. I don't see how the cable could be bad, so I don't fucking know...

Steve Koch 05:16, 16 July 2009 (EDT): Discovered the QPD math was only enabled in the piezo ramp module. Thus, the setpoint was not going to work. Implementing it now in the core "START and read scan..." VI.


 * turned off "reverse acting" and data file 47 may have a successful FTC. If so, this could indicate a polarity problem with the detector.  That will be annoying and / or perhaps easy to fix.


 * VC looks promising. At least the light flickers.  Evident that I have a lot of work to do to get these going.  Need to go sleep.