Galvanometer Heating With Fast Scans on MOM

Stryker and Darcy measured the temperature of the X-galvo on MOM#2 as we scanned, on 23 June 2010, using YSI banjo probe on the distal end of the X-galvo, with thermal joint compound. The banjo temperature readings corresponded to apparent temperature of the block in which the galvo was mounted as felt by a finger. According to Rick and to Steve Sequeira of Camridge Technologies, the galvos are safe up to a case temperature of 50 deg, and not beyond. At a case temperature of 50 deg, the coil in the galvos is about 110 deg.

Scan Parameters and Conditions

Scan parameters (IMPORTANT) were X-2.5V Y-2.5V (peak to peak) bidirectional scans at nominal 0.5 mscec/line (0.54 true msec/line) (acq delay 0). Parking was 8,8 V, and deg/volt was 1.

Active water cooling was on continuously, via the Sutter cooling block and our Thermaltake circulator.

Results:

Scans at 2X barely warmed the galvo, from 28 deg to about 33 deg.

FOCUS Scans at 1X (which moved the galvos from +2.5V to -2.5V) caused the temperature to increase quickly:

Time Temperature
0 33.5
1 42
2 45.5
3 47
4 48
5 49

Stopping FOCUS Scan, galvo cooled to 40 deg within 1 min and to 35 within 2 minutes.

A STACK scan at 1X (with 0 um z-steps and 0 Acquisistion delay), which imposed a short break between each frame, raised the temperature much less:

Time (min) Temperature (deg C)
0 32
2 35.5
4 37

Conclusions:

Even at 1X, we can safely scan for up to 5 minutes in FOCUS or Calcium-imaging mode, in which scanning is continuous with no breaks between frames. We should limit ourselves to 5 minutes continuous at this scan amplitude.

Even at 1X we could make STACK scans for as long as we like, essentially forever, because the break between frames within a scan keeps the temperature from ever rising too high.

These conclusions depend critically on the scan frequency and amplitude. Larger amplitudes or faster scans would cause faster heating. We did not explore larger amplitudes because a 200 micron field with a 40X objective seemed to be as large as we would ever want to scan (1024 pixels would mean 0.2 microns per pixel).

--Michael Stryker, 23 June 2010