Proposal for ORB III, technical specification

1. Emitter

Comparison of alternative components

Due to the fact that HP stopped the production of LST 500 diodes we have to look for another transmitter in the OM:

Data	ELED	ELED	MQW laser diode	VCSEL
Sample manufacturer Sample type	Hewlett Packard LST500	Honeywell GFE1300-5478	Mitsubishi FS-17LSD	Honeywell HFE4080-321
center wavelength	1300 nm	1300 nm	1300 nm	850 nm
spectral width (RMS.)	ca. 50 nm	60 nm	ca. 2 nm	0.5 nm
bandwidth/rise time	266 MHz	200 MHz	0.3 ns	6 GHz/0.3 ns
Threshold current	2 mA	?	3 mA (preselected)	3.5 mA @ 25°C
max. drive current	150 mA	150 mA	28 mA	20 mA
slope	0.001 mW/mA	0.001mW/mA	0.07 mW/mA	0.04 mW/mA
opt. power into 62.5/125 mu multimode fiber min.	0.1 mW	0.07 mw	20 mW	0.5 mW
MTTF	?	?	10^8 h	10^7 h
data sheet	?	GFE1300-547	?	HFE4080-32x
cost	80 US$	?	120 US	60 US

ELED:

+ cheap
+ the wavelength 1300 nm guarantees a low distortion even with a spectral width of 50 nm
- low optical power --> need for high amplification in the ORB, low signal to noise ratio

MQW laser diode:

+ highest optical power
+ a channel could survive even with a fiber or underwater connector damage
+ linearity
- expensive
- calibration at room temperature is not possible, slope and threshold current are temperature dependent

VCSEL:

+ low drive currents --> lower PMT amplification (aging!)
+ linearity
- the wavelength 850 nm results in a slightly higher attenuation
- to measure: can we live with 850 nm and a spectral width of 0.5 nm concerning dispersion

Conclusion:

If we look for replacement of the LST500 diode we have to find laser diodes with a threshold current as low as 2 mA!
The required amplification in the ORB depends on the used transmitter.

2. Fiber and fiber optical components

Three cenarios:

Component	attenuation @ 850 nm VCSEL	attenuation @ 1300 nm laser	attenuation @ 1300 nm with a ELED
Emitter slope	0.040 mW/mA	0.070 mW/mA	0.001 mW/mA
Splitter 90:10?	0.5 dB	0.5 dB	0.5 dB
penetrator	6 dB	6 dB	6 dB
Fiber 125/62.5 µm	6 dB = 2 km * 3 dB/km	3 dB = 2 km * 3 dB/km	3 dB = 2 km * 3 dB/km
ST-Connectors	1 dB	1 dB	1 dB
Receiver	0.5 A/W	0.7 A/W	0.7 A/W
''efficiency''	0.001 A/A	0.005 A/A	0,00007 A/A
Transimpedance of the amplifier converting photo current into voltage	1 k	1k	1k
This is the signal you get at the output of the optical transmission: 1 pe = 50 mV @ 50 Ohms = 1 mA	1 mV (transimpedance of HFBR-2316 receiver is 20 kohms --> 20 mV)	5 mV (transimpedance of HFBR-2316 receiver is 20 kohms --> 100 mV)	70 µV (transimpedance of HFBR-2316 receiver is 20 kohms --> 1.4 mV)

Conclusion: Using a laser diode gives at least 12 dB more dynamic range

3. Receiver module ORB III

Remote control:

based on a Siemens C515 micro controller with integrated CAN interface
2 DACs per channel (amplification discriminator)
LED Display for channel number, amplification and discriminator threshold (if required)

+ setting are reproducible and can be stored
+ calibration can be done program controlled (> 300 channels)
+ C515 micro controllers is well known (same as on the computer Farm control board), evaluation tools are available
+ CAN is an established standard in AMANDA - no additional controllers are required
+ 4 knob control occupies less area on the front panel, easy handling and operation
+ less mechanical components

Optical receiver:

PIN diode with pre amp
voltage controlled amplifier (e. g. OPA 660: OTA with voltage controlled transimpedance)
delay line (if required)
output driver
discriminator (TOT), ECL-output on the back panel

Optical components

E2000 fiber connectors (works with lose tube cable, problem is petrolat)

There are hybrid adapters available: ST - E2000. We can equip the underwater cable with ST's as we have done it before. They were not touched and moved very often, so we can live with the drawbacks of ST's.
The patch cabling should be done with E2000.

Mechanics

Single wide NIM-module
Front panel:

LED Display
control knobs
12 x E2000
24 LEMO's

Back panel:

power supply
flat ribbon connector (discriminator outputs)

Cost estimation (1 US$ = 1,75 DM)

Transmitter	VCSEL or	50 US$
	MQW laser or	120 US$
	ELED	68 US$
Receiver	12 x Pin + Pre amp + pigtail + E2000	70 US$
	24 x LEMO	8 US$
	12 x delay lines	40 US$
	1 x PCB (incl. design, contains 12 channels)	400 US$
	12 x set amplifier	30 US$
	12 x sets peanuts	10 US$
	--->	210 US$ per channel

Final Decisions ?????

Do we really need a delayed output? --> it would save a lot of bugs if not
How many undelayed outputs to we need?
Do we need a discriminator, if yes. what mode (constant fraction, TOT)
What type of transmitter (laser, VCSEL, diode) we are going to use?
Is a different case (VME e.g.) required?
What is the highest optical fiber attenuation the system should be able to deal with?

Status

2 VCSEL are ordered, deep temperature test and linearity test will be done within 2 weeks
the voltage control amplifier is under test, hand wired prototype of receiver chain will be available for further evaluation within 4 weeks
micro controller circuit is under design (schematic capture)