Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Confluence hub for a ton of useful pages: Lasers & Timing Controls
See also: RF locking how-to basics

Overview

(diagram from 20221006)

Image Added


A few explanatory notes

  • The box at the top represents the event receiver that generates triggers based on event codes. The physical device is located in the back side of rack R64A with 12 labeled trigger outputs.
    Image Added

    The control screen for this unit can be launched from the MEC Home screen by clicking "EVR" and then "New MEC LAS EVR". The first six channels are for the MEC short-pulse laser system, and the last six channels are for the MEC long-pulse laser system. The purpose of each channel is labeled in the Description field. 
    Image Added

  • There is a pacemaker for the 120Hz input of the Legend (also passed along to the MPA1 EVOs) and for the 5Hz input of the Gaia. For historical reasons, the pacemaker is kept inside a lockbox in the back corner of the lab next to rack S62 on the floor. The 5Hz channel has been modified to be wider than typical when the input has been disrupted – this is to allow it to act as an inhibit pulse as well as a trigger pulse, which may be a future change to the triggering system.

  • You may notice that the Pulse Slicer is labeled as having two input triggers – one from the "Pulse Slicer" DG645 and one from the "MPA1/EvoHE" DG645. This is for the purposes of the "mousetrap", which looks for light back reflected from the target area. This pulse slicer, located between the first and second multi-pass amplifiers, is a Coherent dual cell-style Pockels cell which enables a single edge (as opposed to an up-down driver, which can open and then close the electro-optic "gate"). It is aligned to passively pass light rather than blocking it – in other words, the polarizers aren't crossed on either side of the cell but rather there is a vertical polarizer on both sides that pass the incoming vertically-polarizer beam. When it's desired to prevent the output of MPA1 from seeding MPA2, the Pockels cell needs to be triggered before the seed pulse arrives so that its polarization gets rotated and the output polarizer dumps the beam. When it's desired to seed MPA2 and then close the Pockels cell late in order to prevent back reflections from propagating upstream, the Pockels cell needs to be triggered after the seed pulse arrives but before the back reflections arrive from target (~100ns window). This is why there are two triggers: when triggering early, the MPA1 DG645's channel AB sends a trigger to fire the cell (i.e. to prevent seed light from entering MPA2); when triggering late, the Pulse Slicer DG645's channel CD sends an inhibit signal to the MPA1 DG645 to stop the early trigger from firing and channel AB sends a trigger delayed by several ns to fire the cell (i.e. after the seed pulse has gone through but before any back reflections return). In all these cases, the exact delays and the inhibit settings must be set correctly.

  • There are two UNIBLITZ shutters used for the short-pulse laser system which need synchronized triggers. Because these shutters are mechanical, the actuation mechanism is relatively slow (~few ms for the 6mm shutter and >100ms for the 65mm shutter). Because we want the shutters to be fully open by the time the laser pulse arrives, the "open" command for the shutters must be sent well in advance of the triggering of the laser itself. This is done using the event sequencer and a dedicated event code 177. The proper event sequencing is shown in the top right corner of the timing diagram. If other steps are added to the event sequencer, the important point is that there need to be a total of 12 beam deltas of delay (using a sync marker of 120Hz) between the 177 event code (commanding the shutter to open) and the 176 event code (associated with the firing of the laser pulse itself).

  • Because the UNIBLITZ shutter tends to be used only when the laser is firing on-target, much of the time it is desirable to hold the shutters open and to prevent the shutters from firing. To hold the shutters open, change the polarity of the AB and CD output channels of the UNIBLITZ DG645 as shown above, with POS holding the shutters open and NEG holding the shutters closed. To avoid triggering the shutters while performing alignment, be sure to also inhibit the triggering of these channels as well (or else you could accidentally trigger the shutter to close at exactly the time the laser is being fired (big grin)). Below shows the UNIBLITZ DG645 configured to inhibit triggering of the shutters (i.e. by choosing "AB,CD" under "Trigger Inhibit") and to hold the shutters open (i.e. by setting the polarity of both AB and CD to "POS")
    Image Added

  • There are also times when this single shot mode is important for triggering cameras to look at single-shot images correlating to a shot on-target, e.g. for beam profile, beam pointing, back reflection detection, etc. This is accomplished using channel EF of the UNIBLITZ DG645, which serves as an inhibit source for the GigE DG535 box (which has no IOC or network connection). This is also controlled using the polarity: if you don't want to inhibit the GigE triggering at all (i.e. you want it to just trigger at the nominal 5Hz repetition rate), then the polarity of channel EF of the UNIBLITZ DG645 should be set to NEG; if you do want to inhibit the GigE triggering (i.e. you want to only uninhibit/trigger the GigE cameras upon firing of a single shot with the event sequencer, like the UNIBLITZ shutters above), then the polarity of channel EF of the UNIBLITZ DG645 should be set to POS. This could all be controlled by a Python script if we wanted to make a Jira ticket out of this. (Side note: you may notice some instructions about special triggering cases. If you want to inhibit the GigE cameras to run at e.g. 1Hz but you don't want to use the event sequencer for whatever reason, then the GigE shutters can be inhibited, the event code input for the DG645 can be changed to a 1Hz event code, but then the timing of the GigE inhibit needs to be altered in order to account for the fact that you aren't using beam deltas in the event sequencer anymore so it needs to be added explicitly. It's very rare that people want to do this, so consider this more of a warning that there's something more you need to consider if you're trying to do something without the event sequencer.)


If you have questions, please make a comment and I'll add details!Image Removed