Two-photon: further information

How does the system work?

Why do we have a home-made 2 Photon?

• It is substantially cheaper.
• It is more customizable - the system is easier than many modern systems for us to modify. We also have access to the control software.
• It is better to have a dedicated 2P system than to try and squeeze the feature onto one of our heavily used confocals. This would also make the 1-photon confocal systems more difficult to use for people not interested in 2PE.
• The commercial range of systems is somewhat limited by a patent now held by Zeiss, this expires very soon . . .

Does the system have any limitations?

Yes.

• It is not as automated and easy to use as an off-the-shelf system.
• It is not particularly fast
• Limited to 1024*1024 acquisition
• Changing wavelengths requires some manual alignment

Current bugs and work in-progress

Zoom

Distortion compared to no averaging. z1 is ok. All others zoom and the y axis is compressed, more zoom more compression. X-axis zooms succesfully.

ROI zoom/specifed beam parking needs to be added for ablation/uncaging.

Time-lapse

Time-lapse capability needs to be added

Averaging

Very slight image shift between unaveraged and averaged image, not a particular problem though.

Ideally needs to work at all scan speeds (currently 512lines 2second scan or slower, 1024lines 4 second scan or slower). This produces some speed limitations.

In general the machine is quite slow. 4 second 1024 scan with no averaging = 5 seconds, same with averaging 4 = 43 seconds averaging 8 = 69 seconds. 8second scan av2 = 35 seconds, 8seconds av4 = 68 seconds

AOM

AOM control does not work and has been replaced by an attenuation wheel and manual shutter. Works very well but not as convenient as computer control.

Detector Noise

PMT2 (NDD straight) is noiser than NDD PMT3. In dark with detectors shut off, b9515, gains that produce similar noise level were 652 and 930 respectively. Better cooling?

Alignment

Alignment - seems quite unstable. Is pointing stability for precomp within spec? Consider motorized actuator for long term.

With BE at most focused (255), the path is not clipped. Expanding the beam - the beam is cliped around 200, the 10x objective is only well filled at around 180. The shape of the scanned beam at the back aperture of the objective is not gausian. Nor is it after the objective (ie card held about 1cm above 10x objective). Nor is 633, the path to which it is aligned.

Multi-channel imaging

Display seems to strugle with 2 or more windows. Jittery updating of windows (ie not smooth scan of both)

Not able to image 2-channels over a Z-stack

Depth adjustment

Would be good to be able to alter laser power and/or gain over z