Blog

• Issues observed by UED team in previous operation;

  • Present firmware does not run in the LCLS-II DAQ.  Needs update to PGP4? Action:  Dan to speak with Matt and determine whether an update is possible in time. (Incl. Zach, Chris, Silke).
  • Verify proper function in Low Gain mode.

• Hardware status for camera, including software and firmware

  • Action:  TID to visit UED to read out firmware version (Matt)

• Testplan:

  • Goals

Making note of which MPOD channels correspond to which detector (ePix10k small/Quad/2M, ePix100, others) and what the operating voltages are. 

ePix10ka small OR ePix100

*Might vary based on temperature sensor reading. Typically ranges from 2-4 V (as long as temperature stays between 3-10 C). 

ePix10k2M

Table entries correspond to Q0/1/2/3

** Turn all AVDD on first, wait 5-10 seconds, then turn DVDD on. (Kaz said it ensure 2M configures properly?) 

This blog post is where I'll document the process of setting a gain mode for the epix10ka small (no.1) and taking pedestals.

(How to create a new gain mode text file to upload)

10k small set up and powered up in the Marietta. Humidity and temp stabilizing around 5% and 8.9 deg C. First gain mode manually configured is AHL Forced L: tr_bit = 1 and mode_4.txt loaded in (pixel map is a brick red or brimstone red) → run 27. Second gain mode configured is AML forced L: tr_bit = 0 and mode_4.txt (pixel map still same brick red color) → run 28.

Both configs tested just in case something would be different between the two, not that there's any reason to assume it's different.

Pixel map:

Pedestals:

Next step: compare this to Fixed Low array, map the difference array.

Setup:  RIXS-CCD in test chamber.  Turbopump on, Pirani reads 2.7e-8 with Polycold running.  Foil covering window to block room light.  Archon configured for binning mode, external trigger.  Frame rate set by burst period setting of Silent Waveform gnereator.  Configured according to RIXS-CCD procedure at https://drive.google.com/file/d/18SGIGRSk_v71dKv3i-ESIEPfAwkLIDIv/view?usp=sharing

Method: Define noise stats box to include a block of pixels avoiding the non-pixel gaps in the display.  Since the block contains sequestial reads of the binned pixels, the standard deviation over the block is a ~proxy for noise.  Record std. dev and absolute singal at 50 Hz, 100Hz, and 200Hz to measure leakage.  Gradually increase temp from -120C, pausing 15min at each temp to stabilize before recording.  Save 10 frames at each setting for later analysis.

Surprising to me that the noise and leakage never really seemed to take off until well above -75C....expected to see it increase by x2 per 7K or some such, but I guess that's only around room T.  Will need to look at the data to understand how I'm confusing myself.

Wondering whether something had just railed, I poked a small hole in the foil over the window, and found the detector was still working more or less as expected even at -50C.  Decided to take it all the way to 0C, expecting it to become unusable at some point.  Not sure that ever happened, but at least the noise started to go up

Went back through the data using a script written by Kaz that actually calculates the noise for the entire device.  I believe it includes unit cells that don't have active pixels and may thus underestimate the noise.  The results differ in detail but the general trend is similar to the 'box' proxy noise estimate.  The noise from the script is plotted against the TempB reading.

Using 3.3e-/DN, the noise at -61C (TempB) is still below 9e-.  Can't assume that the device temperature tracks TempA or TempB well.  

The new 2M has bee scanned through all modes and ranges using the internal pulser, thanks to firmware updates by Maciej and some hand-holding from Dan Damiani.

The calibration scan is run with the command epix10ka2m_offset_calibration -p 1 -D DetLab -I 0 -i 0 -v 0 -V 1 -e 4608 -s 7.  If you type epix10ka2m_offset_calibration -h you can get a description of the switches.  That script works in ASC...the equivalent script in MFX has a different name and is not working well as yet).

We've been able to get enough flux directly from the Mini-X to flip the autoranging circuitry for the central part of the image.  We've also aimed the Mini-X at a copper target to measure Cu fluorescence flat fields.

Beam time 8/10/20:

• Detector in helium, set bias to 120V per MFX.
• Water jet observed to cause ring, not strong enough to cause auto switch.
• Ad lysozyme to jet to up signal....looks good, not sure whether we saw switching or not.
• Vent He, place metallic glass sample in beam.
• Plan:
  • Record biases and currentsTake 5-pedestal sequence
  • Use metallic glass target (which hopefully will not melt before giving high flux)
  • Run attenuator scan in fixed low gain, witness via Wave8
  • Repeat in AHL
  • Repeat in AML

Presently inviting comment.  In review by LCLS management, to define the requirements governing the VFCCD/ePixM downselect.

https://drive.google.com/file/d/1PgvG7wCCWZ7H-8hWNvQhxrbv49bSvtM9/view?usp=sharing

Two Projects were presented and approved at the March 15 PEC meeting.  One covers the fab of one new in-air copy of the ePixBig plus four new quads for MEC.  The other project re-baselines ePixM and aims it toward a downselect for the SXR Imager deliverable for NEH 2.2.  The PEC presentations have been uploaded to the Detector Group GDirive for reference. 

Now posted at https://docs.google.com/document/d/1wAvTAqMOwkrs6yxerK5UrQ6Lx2x5GeBdfGIM4rcXMOI/edit#heading=h.nrnw03t7conb

The TID-AIR folks have prepared a briefing describing the persistence artifacts observed recently in ePix10k, where they come from, and propose dfixes.  Upshot is:  A firmware fix to induce a self-triggered reset frame has been implemented, which resolves the issue in a way that's transparent to users.  A hardware fix has also been implemented in the design for all future ASIC fabrications.  Read about it here:  

https://drive.google.com/file/d/1OHfW9YDoEERSSeKgS_8petTpXx-Zlh31/view?usp=sharing