Pending DOM Changes
Draft
Favored changes.
Component Changes
- Changes to filter circuit between DC-DC converter and
communications front end that allow metalized polypropylene film
capacitors to replace aluminum electrolytic capacitors.Rev3
- Resistor footprint in parallel with DOM On-Board LED.
This will provide control of the light pulse trailing edge.Rev3
- Remove R78 on Sheet 6. It duplicates R377 on Sheet 7.Rev3
- Communications transformer, T5, will be replaced with a suitably
reliable product with the proper turns ratio will replace the
ill-suited TT1.5-6 transformer from Mini-Circuits. (Candidates)
(dual footprint until parts available?)Rev3
- 1:2 impedance ratio transformers will replace T2 and T3 on the
Local Coincidance board. (Candidates) (dual footprint
until parts available?) Rev4?
- 1:1 ratio transformer will replace T8 on Sheet 10. In Rev 1
and 2, the
transformer was not physically robust. It will be replaced by a
suitably reliable transformer. (Candidates) (dual
footprint until parts available?)
- Ferrite Beads B23 and B27, at the input to the DC-DC converter,
on the Power Supply sheet, become Delevan SMB2.5-1 Surface Mount
Shield Bead..Rev3
- L7 and L8 on sheet 14, the oscillator page. Delevan looks
most promising. (Candidates)Rev3
- Change 220 uH inductors L5, L11, L13, L15 from JW Miller part to
one from an approved vendor. (Candidates) .
(change on hold 6/20/03)Rev4
- Change 10 uH inductors L1, L6, L9 from JW Miller part to one from
an approved vendor. (Candidates)
. (change on hold 6/20/03)Rev4
- U127 should instead be a JP ref. designation, since it is a
connector for an off-board cable.
- Additional power and ground pins of the EPXA4 must be picked up
in this revision. It includes power for PLLs,
I/O banks, and grounds for PLL and I/O banks.Rev3
- Move Varistor RV1 from power supply sheet, sheet 4, to a position
in parallel with C89 on communications page, sheet 10. This
improves the effectiveness of DOM MB protection from unexpected
transients or blunder.Rev3
- A resistor in series with the grounded end of CR8 should be put
in to control saturation current behavior.
- Two BAS70-04 series diodes should be added from the delay line
input to ground to guaranty clipping of PMT signals below 4 Vbe below
ground.
- C481 needs an "*" to indicate that the part is 'do not load'
because it is only needed for a higher resolution DAC. (change on hold
6/20/03)
- C470 needs an "*" to indicate that the part is 'do not load'
because it is only needed for a higher resolution DAC. (change on hold
6/20/03)
- R87 on Sheet 10 should be 210 ohms, and a footprint size
greater than 0805 is desirable. 1210 recommended . (change
on hold 6/20/03 value TBD)
- R818 on Sheet 10 should be 137 ohms, and a footprint size greater
than 0805 is desirable. 1210 recommended.. (change on hold
6/20/03 value TBD)
- R125 and R137 change value from no-load to 750 ohms in order to
properly match the expected impedance of the neighbor-to-neighbor local
coincidence cable . (change on hold 6/20/03 value TBD)Rev4
- R143, R376, R384, and R387 change from 75.0 ohms to 68.1
ohm. . (change on hold 6/20/03 value TBD)Rev4
- Put a resistor between CPLD pin 18 and EPXA pin G21 to protect
drivers in the case that both pins drive out in opposite directions at
the same time. (omission from Rev 2 changes. Probably not a
problem)Rev3
- Change CPLD from XC2C256_TQ144 to XC2C384_TQ144 logic cell part.
for reliability reasonsRev3
- Create power supply circuit to deliver -3.3V analog supply to
support the
HFA1135 changesRev3
- Create power supply circuit to deliver +6.6V analog supply
to support the
HFA1135 changesRev3
- Change U53 from AD8014 to HFA1135 amplifier, along with the usual
limiting changes.
- Install the option of BAS70-04 diode networks on the inputs
of U44 ,U48, U42, U54, U66 to protect the stages from excessively high
or pathological transient signals from PMT input
- Resize resistors at inputs of U41, U44, U48, U54, and U66 to give
workable diode clipping behavior.
- Change feedback networks to yield gains of x16, x2, and x0.25 in
the ATWD channel 0, channel 1, and channel 2 inputs to properly match
up with input diode clipping.Rev3
- Remove 3.3V power line from analog front end. -Replace with
connection to +5VRev3
- Add COMM_RESET line between FPGA and CPLD including pull-up.Rev3
- U53 will be replaced with a clipping amplifier HFA1135.Rev3
- U50, AD626AR, Put jumper between pin7 and ground to
accommodate control of range (with EPXA4, the output is near the limit
of the ADC span)Rev4
Layout Changes
- Net MEMVIO becomes an island of copper straddling the SD RAM
memory chips and the SDRAM Controller section of the CPU chip.Rev3
- Net MEMCORE becomes an island of copper encompassing both SD RAM
Memory Footprints.Rev3
- Net VddC0 becomes an island of copper covering the various PLL
power pins..Rev3
- The inverting input of all high-speed amplifiers (AD8005, AD8014,
HFA1135, AD830 and AD8130), make sure the components connected to
the inverting input are as close as possible to the inverting input.
- For all high-speed amplifiers (AD8005, AD8014, HFA1135,
AD830 and AD8130) have inner planes carved away from the
inverting input pins and the connected pad of any component connected
to the inverting input. 2mm clearance needed around package.Rev3
- Make sure that the end of the feedback resistor connecting to the
non-inverting input is under the non-inverting input pad of any high
speed amplifiers. (Particularly avoid having the either pad of
the feedback resistor underneath the non-inverting input.)Rev3
- Special footprint for JP5, the SMM-105-02-S-D-LC-K, which
has pass-through holes plated through, and connected to SMD pads for
the basic connector.
- Routing of pulser transformer leads away from front end pulser
switching signal traces. series resistors will help.
- SD_nCLK trace length to match length of other lines going to
memory components.
- Route ATWD input signal traces away from or screen from ATWD
digital signal traces.
- Per Albrecht Karle's request, locate DS1 close to the edge of the
board. (as close as possible, with leads parallel to a tangent to
the board.
- If possible, put +6.6V and -3.3V power on planes extending into
anlalog front end.
Possible changes
- The ferrite beads, B19 and B30 on U44 pin 4 and pin 7 may be
replaced with
direct connection to power planes, if bench tests prove the change is
justified. The signal induced noise on the power pins of U44
appears to be insignificant at 1% of the output signal. .
- Some improvement to get the 'ringing' out of the pulser.
- Change from 100V to ±5V DC-DC converter to a 100V to +5V
DC-DC converter plus a +5V to -5V switcher on the DOM MB. (not
favored by Bob)
Candidates for T5
Rhombus T-10207 or the T-13633.(but unfortunately Rhombus isn't QML,
nor do they provide the necessary reliability informaton)
Vanguard
Delevan
Datatronics PT-31406 http://www.datatronics.com/pdf/adsl_pt314.pdf
Pulse Engineering B2105 http://www.pulseeng.com/pdf/B950.pdf
(except for temperature range)
Bourns
Beta Transformer MLP-2012 http://www.bttc-beta.com/download/download.asp?Type=MIL-STD-1553&File=bumlp2000ds.pdf
Candidates for T8
Vanguard MUB55D http://www.ve1.com/MKMsers.htm%00
Delevan
Pulse Engineering CX2041 but looks rather fragile
Pico Electronics
Candidates for T2, T3
Vanguard MUB510D http://www.ve1.com/MKMsers.htm%00
Delevan
Pulse Engineering CX2045 CX2045L but both look too fragile
Pico Electronics
Candidates for L7, L8
Vanguard 33031 or 33031 NT 80 ppm/ºC
Delevan 1330-34K http://rust.lbl.gov/~gtp/DOM/dataSheets/delevan_1330.pdf
Candidates for 220uH inductor
Delevan 5142-224K http://www.delevan.com/PDF_DOCS/5142.pdf
Pulse Engineering P0770-224h http://www.pulseeng.com/pdf/PC500_1.pdf
Candidates for 10uH inductor
Delevan 5142-103K http://www.pulseeng.com/pdf/PC500_1.pdf
Pulse Engineering
http://www.pulseeng.com/pdf/PC500_1.pdf
P0770-103