Make a BPM out of a Damper - PowerPoint PPT Presentation

1 / 10
About This Presentation
Title:

Make a BPM out of a Damper

Description:

Resistive Wall Monitor. Broadband Cavity. FAST. DACs 27 MHz. FAST. DACs. Power. Amp. Transverse ... Noticeable, timing must sync to accelerator clock. ... – PowerPoint PPT presentation

Number of Views:29
Avg rating:3.0/5.0
Slides: 11
Provided by: vincepa
Category:
Tags: bpm | clocks | damper | make | out | wall

less

Transcript and Presenter's Notes

Title: Make a BPM out of a Damper


1
Make a BPM out of a Damper?
2
Digital Damper
53 MHz, TCLK, MDAT,...
106 / 212 MHz
Stripline Pickup
FAST ADC
Monster FPGA(s)
Minimal Analog Filter
14
Transverse Dampers Identical X Y
FAST ADC
Minimal Analog Filter
Stripline Kicker
Power Amp
Ethernet
FAST DACs
2-10
gt 27 MHz
Resistive Wall Monitor
FAST ADC
Minimal Analog Filter
Longi- tudinal (Z) Damper
Broadband Cavity
Power Amp
FAST DACs
2-10
3
Damper -gt BPM
Stripline Pickup
FAST ADC
Analog Filter
M E M O R Y
12
FAST ADC
Analog Filter
BPMs H V
VME
Stripline Pickup
FPGA
FAST ADC
Analog Filter
12
FAST ADC
Analog Filter
Delay line per ADC
FPGA
212 MHz
Timing Logic
53 MHz, TCLK, MDAT,...
4
BPM Pick-ups
V M E
FADC 4 Chan/2 BPM
Sample Timing
UCD
Clocks and Timing
VME CPU EDB PMC
ACNET
BLM
5
BPM Pick-ups
V M E
FADC 4 Chan/2 BPM
Timing and EDB/BLM
BLM
Clocks and Timing
VME CPU
ACNET
6
So What Do We Get?
  • Minimum A Sample every bucket, allowing
    filtering and averaging everything with
    everything. Multiple data streams.
  • Synchronous sampling and phasing can be beam
    triggered.
  • Sampling based on the beam structure, i.e the
    FPGA data rate is proportional to the beam
    structure.
  • We can time-separate the proton signal from the
    antiProton signal.

7
FPGA Logic example
Memory Interface
Analog Scope
Out-of-bunch Data
Memory Interface
Threshold
Anti-Gate
212MHz Samples
Memory Interface
Turn Average
Gate
Memory
Memory Interface
Bunch n TbyT Filter
Bunch Structure Timing
Memory Interface
Bunch m TbyT Filter
8
What is the bad news?
  • It isn't a perfect fit to our problem but the
    core design is close enough that we should be
    able to use it.
  • It isn't a production board just a good core
    design we can build upon.
  • It is complete over-kill but it costs 400-500
    per channel today, quantity one.

9
What needs to be done?
  • Determine the Analog Filter to use
  • Nothing??
  • Low Pass Filter only?
  • Switchable Bandpass Filter to spread the pulse
    width on uncoalesed bunches?
  • Determine functionality of current board and
    stuff another one for our testing.
  • Soon, channel decision. Then layout new board
    and simulation of VHDL logic.

10
Webber Questions
  • are there parallel, simultaneous position
    processing issues in the new requirements that
    the FDB architecture can/cannot satisfy?
  • Nope
  • what are the requirements for switching between
    modes in a short time?
  • Run several modes in parallel
  • is the FDB pipeline/FIFO processing
    architecture limiting in any significant way
    relative to TeV requirements?
  • No FIFO, FPGA pipeline runs faster than data
    rate.
  • are implied VME backplane data rates
    acceptable?
  • VME rates are ACNET request dependent. Boards can
    buffer data.
  • what might be possibility of excessive loading
    on VME CPU?
  • Programmer dependent.
  • what are the triggering implications and
    requirements on external timing/triggering
    system?
  • Noticeable, timing must sync to accelerator clock.
Write a Comment
User Comments (0)
About PowerShow.com