Picometrix
Picometrix | ||||
---|---|---|---|---|
| ||||
Description: | Picometrix | |||
Manufacturer: | Picometrix | |||
Model Number: | T-ray 2000 | |||
Quantity: | 1 | |||
Location: | Workstation 'USA' | |||
Department barcode number: | EExxxxxx
|
Contents
- 1 Presentation
- 2 Manufacturer's Data
- 3 Alignment Notes
- 4 Bandwidth and Alignment
- 5 How to align T-ray heads
- 6 Control Unit Alignment
- 7 Fiber Coupling
- 8 OFR maintenance
- 9 Fiber care and repair
- 10 Wet cleaning of optics
- 11 Picometrix Refractive Optics
- 12 Picometrix Tx heads
- 13 Picometrix Rx heads
- 14 Software
- 15 Problem with translation stage
- 16 Optimizing Bandwidth
- 17 Cleaning up the Time-Domain signal
- 18 See Also
- 19 Back
Presentation
Manufacturer's Data
- [2] GDC photos
- [3] GDC Alignment
- [4] T Ray 2000 Hardware Manual Rev.1a
- [5] Additional comments on GDC Alignment
Alignment Notes
The first stage of the GDC is a Faraday Isolator assembly. This consists of a rotator, polarizer (internal), Faraday rotator crystal, adjustable rotator (as attenuator) with external knob, and a final polarizer (cube).
The damage threshold for this port is 400 mW. The maximum power into this port is 300 mW.
There is a large vertical knob on the GDC. This is the adjustable rotator that attenuates overall transmission through the assembly. At minimal attenuation, the transmission through the assembly is ~10%. If your best alignment transmission is a factor of 2 or more less, you may have a mis-alignment with the transmission due to internal-reflections in the Faraday rotator crystal. This may work with lower attenuation, but the alignment will drift significantly over time. Therefore, please be sure to correctly align the beam into the input assembly.
There is an adjustable iris before the Faraday assembly. This is useful as an alignment aid, however, the beam does not quite passes through its center.
Before commencing alignment, make sure that the fibre is disconnected from the output port. The beam will need to be attenuated to 20 mW before the final alignment is undertaken. This is very important to avoid damage to the fiber tip. These are special APC PM fibres that are polarization maintaining, and cut to specific lengths. They are easily damaged and expensive to replace.
Bandwidth and Alignment
The optical alignment of antennae independently affects both the signal amplitude as well as the bandwidth. The bandwidth effect is due to the different spectral components emanating from different parts of the antenna geometry. The higher frequency components are emitted from the center of the bowtie antenna. A similar effect occurs with the receiver. Whereas the transmitter optical alignment is fixed inside the head, the alignment between the receiver bowtie antenna and the T-ray depends on the overall system alignment.
How to align T-ray heads
Use a long ruler to iteratively align up the head bodies of Rx and Tx modules in the X-Y plane by maximizing the signal. Loosen the 4 screws holding the lens mount, and iteratively adjust the receiver lens X,Y position to optimize the bandwidth slope. You can use a previously captured waveform as a reference, and make the slope as shallow as possible to optimize bandwidth independently of signal amplitude.
Control Unit Alignment
There are two fiberports in the control usnit. One launches the light into a polarizing beamsplitter cube and into the retroreflector. Avoid adjusting this beam path!. If the Rx head power is low, check the coupling efficiency of the fiberport "B", and carefully adjust the plunger screws if necessary. Do not touch the XY screws on any fiberports.
Fiber Coupling
Coupling light into a single-mode fiber is no easy fete. The physics of are as difficult as the mechanics. The theoretical maximum efficiency is around 70%, however, I have never seen anything much better than 60%. The Picometrix fiber is APC (angled contact) to reduce Fabry-Perot interference. This makes it even more difficult. Today I got an efficiency of 53%, however, I expect alignment will drift. The two adjustment knobs on top of the GDC are there to trim out this drift. If you are experiencing drift problems, the most likely cause is dirty or damaged fiber tips.
OFR maintenance
The Fiberport can get seized up. If this occurs, disassemble the unit and polish the mating surfaces with 12-micron grit Alumina sheet. Once reassembled, set the XY position to center the lens (see below). This applies to all versions of OFR, with APC fiber. After the celns is centers, never adjust the XY screws. Turn the plunger screws all the way in, and then release 2.5 turns. This is a good starting point for further alignment.
There are 2 kinds of fiberports used in the Picometrix. The PF-X-11 has an angled shaft, and the correct X-Y position for the lens is in the center. The PF-X-7 is for a smaller beam diameter, and has a straight shaft. It is more difficult to align. The correct XY position of its lens is off-center. Once the XY position is correct, don't move the X-Y screws. Only adjust the plunger screws.
Fiber care and repair
To clean, gently press and stoke the tip against soft dry (or wet) cleaning paper. The Picometrix comes with "Diamond" brand fibers with metal sleeves. If damaged, these cannot be polished because the metal wears down slower than the glass core. Conventional (Thorlabs) fibers can be polished and repaired using special equipment, at 8'(APC), starting with 12 micron grit, and going down to 1 micron or less. We do not have this equipment.
Wet cleaning of optics
Methanol: Hydrophilic and leaves less residue. Acetone: Hydrophobic. Good for cleaning off oils. Can leave some residues. For 2-step cleaning, use acetone followed by Methanol. For one-step cleaning, use Methanol. Best if purchased and stored in glass bottle with glass stopper and pippette.
Picometrix Refractive Optics
There are both Silicon and HDPE polymer lenses available for the Picometrix. The Silicon lenses have lower losses at high frequency, but also suffer much more Fresnel reflection due to their high refractive index. Despite their higher losses, the polymer lenses have been found to offer slightly better overall performance, even at the higher frequencies.
Picometrix Tx heads
There are two heads available with different antennae geometries. The 'Standard Bowtie' head offers a very high signal level at the lower frequencies. The "Snub Bowtie' geometry suppresses low frequency resonance, resulting in a flatter bandwidth. This geometry offers slightly higher bandwidth at the expense of low frequency dynamic range. All Picometrix transmitter heads have a limited life expectancy while powered on with light. Please turn off power, or laser, (or both) when not in use. The output power from the T-ray 2000 heads is comparable with the current model 4000 and 5000 systems. It is very high.
Picometrix Rx heads
The T-ray 2000 GaAs heads perform relatively poorly compared to the newer InGaAs heads (made by Picometrix) that are primarily responsible for the superior performance of the modern T-ray systems. Furthermore, the new InGaAs receivers are compatible with 800 nm, however, they are not available for sale. Picometrix only wants to sell new systems. Unlike transmitters, the receivers are very reliable.
Software
Sensor calibration file is C:/Windows/LBDAS.INI Make sure to back it up in addition to the installed directory.
Problem with translation stage
An intermittent problem observed is the Picometrix translation stage getting stuck between its inner limit switches. To resolve the lockout, simply rotate the translation screw a few turns, and then perform a software XY reset.
Optimizing Bandwidth
Fiber power: Due to Self-Phase-Modulation, the optical power into each head should be less than 2 mW for optimum THz bandwidth. The GDC pre-compensates for the fiber dispersion by adding dispersion and stretching out the pulse. As the pulse travels through the fiber, dispersion in glass causes the pulse to re-compress. The maximum peak pulse power occurs right near the end of each of the fibers at the Rx and Tx heads. This peak power causes another broadening mechanism; Self-Phase-Modulation (SPM) to once again re-broaden the pulse. The SPM strongly depends on the peak power. At this point in the system, the temporal duration of the original laser pulse is important. Shorter pulse is better.
GDC compensation range: The GDC was designed to accommodate fiber lengths up to 6 m. Our system only has 3 m of fiber. This means that we have ample compensation available to accommodate short pulse lasers like the Mira. see: https://www.newport.com/The-Effect-of-Dispersion-on-Ultrashort-Pulses/602091/1033/content.aspx
Cleaning up the Time-Domain signal
Multiple peaks in the signal can reduce bandwidth and can be caused by incorrect waveplate alignment. The Picometrix uses polarization maintainig fibers that work by inducing a birefringence in the fiber. This means that light launched with one polarization, will travel at a different speed compared to that launched with a different polarization. The waveplates at each fiber port ensure that all the optical energy is confined to a single polarization relative to the fiber geometry. If you see two peaks in the time domain signal, it means that there may be incorrect waveplate alignment somewhere in the system because the pulse is split, taking different times to traverse some length of fiber. Rotate the misaligned waveplate to smooth out the time-domain signal.