Status: HFT Readout Electronics

1
Status HFT Readout Electronics

• Xiangming Sun

• Andrew Rose
• Chinh Vu
• Howard Wieman
• Leo Greiner
• Michal Szelezniak
• Thorsten Stezelberger
• Xiangming Sun

2
Motivation

• RHIC physics requires thin fast detector near
  the collision point
• Thin Si-detector (with Monolithic Active Pixel
  Sensor ) is one of the solutions
• New detector needs new readout system

3
Outline

• 1) Introduction the readout electronics
• 2) Design, Performance and Discussions3)
  Summary and/or Outlook

4
Electronics Setup the Chip
Mimostar2 sensor chip, produces analog signal
which carries particle information passing
through it 660128 pixels, 50MHZ output
Daughter card which has ADC, SRAM and FPGA on it.
ADC will digitize the analog signal form
Mimostar2 8 channels ADC,12 bits/channel
5
Electronics Setup - Motherboard
Motherboard which provides power, clock, and
trigger for Daughter card
Microtonix board can collect data from multiple
Daughter cards and send it to optical fiber
(large distance transmission).
6
Electronics Setup - SIU
SIU( Source Interface Unit) is a part of DDL(
Detector Data Link ) which can transfer data
through long distance (a few hundred meters) with
high speed 2G bits/sec This device is
compatible with STAR daq1000
7
Electronics Connections
Linux PC
8
Firmware Structure
9
Question

• When we think we have implemented these
  functions( Fullframe, normal mode and hot pixel )
• How do we test its functionalities?

10
Spike pixels
We noticed that some chips on flexible cables do
not behave as stable as expected. There is a
number of pixels that seem to have DC level
significantly different from the majority of
pixels. This distortion seems to be related to
the middle of a frame. This is still
under investigation
Start frame
end frame
11
Fullframe vs. Normal

• The idea is
• Regarding these Spike pixels as clusters, one
  should get some addresses from normal mode that
  are comparable with Fullframe ADC data.

12
Preliminary Results
13
Hot pixels

• The idea is
• Masking all the pixels in chip except some
  pattern, then flushing the chip by LED, one
  should reconstruct the pattern back by the
  address got from normal channel.

14
Upper is the picture we draw, blue part is dummy
region of the chip. Only red part is not masked
by hot pixel Left is the picture we reconstruct
by addresses in the normal mode
15
Discussions

• The correctness of fullframe mode has been tested
  previously by other means.
• All three work modes are tested
• The system works !!!

16
Outlook

• Test the stability of the system.
• Add warning feature to system for the case of
  errors.
• Beam test at ALS.