May 7, 2001 M. K. Brewer R. Millan-Gabet Fringe Scanning Test using AO board + Interrupts ================================================ This seemed to go well. We attached a Wavetek to the GPS clock interrupt and installed an interrupt handler to ramp the piezo. We used this to scan a laser interference pattern and looked at the result on an oscilloscope. The result looked like a good sine wave. Since we are still unsure if ramping the piezo with the DAC is going to produce acceptable data, I would strongly recommend a full up test of this in June. That way, if it doesn't work well, you will have time to order a programmable waveform generator and get it up and running by the fall. Some additional details on test described above ----------------------------------------------- * piezo used: "visible table" piezo, mount & driving electronics * light source: red HeNe laser * detector: photodiode * interferometer: visible table one, set up in double pass * CPU is interrupted by external signal (TTL from Wavetek) and in response advances piezo position by incrementing the voltage in a linear ramp using the Matrix AO board We wanted to simulate the following situation: 256 data points/scan use full piezo stroke (30 um mechanical) sample at 5 points/fringe rate of several scans per second Therefore: total OPD range: 4 x 30 um = 120 um (because of double pass interferometer) required step size: lambda/5 = 0.6 um/5 = 0.12 um/step or 5V/120um * 0.12 um = 5 mV/step: increment programmed in VxWorks code for output via AO board every time an interrupt is received. interrupt rate: 2KHz => 0.5 ms/step * 256 points/scan = 0.128 s/scan or 7.8 scans/s. measurement on photodiode output seen on oscilloscope validated all above nominal parameters. fringes looked smooth, with no signs of stepping or ringing. Open Question ------------- Will the mechanical response of the real scanning piezo mount, given also the different driving electronics, to be used in the new beam combiner designs be as good? We'll find out in June ...