Background and Mission
The Circuits and Systems group consists of about 20 researchers working at the Electronics Laboratory of the Department of Electrical Engineering at the University of Oulu. Its main activity is in the field of electronic and optoelectronic circuit and system design. The main interest of the group is devoted to certain novel devices, circuit topologies and functional units, although the group is also interested in applications, especially in the field of electronic/optoelectronic measurements.
The main research fields are:
- time-to-digital converters and timing circuits
- generation and detection of powerful and high-speed electrical and optical pulses/transients, and breakdown phenomena in semiconductors in general
- development of pulsed time-of-flight laser range finding and Raman spectrometer technologies, especially for industrial applications
- radio telecommunications, including linearization of power amplifiers, AD/DA conversion and baseband blocks, frequency synthesis
In the following, some details and results of the work of the group are given in selected important research fields.
Time-to-digital converters and timing circuits
A sub-ps-level resolution CMOS time-to-digital converter based on time domain successive approximation interpolation
A time-to-digital converter (TDC) with sub-ps resolution and a ms-level dynamic range is under development . A digital counter with a ms-level dynamic range keeps track of full clock cycles between start and stop timing signals, and interpolators resolve the fractional parts of the clock cycle between the arrival moment of the timing signals and the edges of the clock signal. The sub-ps resolution is achieved with interpolation based on a cyclic time domain successive approximation (CTDSA) method utilizing a pair of digital-to-time converters (DTC) with sub-ps level delay adjustment resolution.
The research work focuses on improving the dynamic range and linearity of the DTC needed in the interpolation method. Since the proposed DTC can reach the ns-level linear dynamic range, the time-to-digital conversion can be performed with a counter and single interpolation stage, without the need for a DLL (delay-locked loop), which simplifies the design of the TDC architecture. A stable, low jitter ~200 MHz clock oscillator could be used as the reference clock.
The larger linear range of ~5 ns of the DTC is achieved by using a differential instead of a single-ended current switch, a two-stage open loop comparator instead of simple two-stage comparator, and a load capacitor matrix. The propagation delay difference between the DTCs, defining the resolution of the TDC, is implemented by digitally controlling the unit load capacitors, and also by digitally controlling the current of the differential current switch discharging the load capacitances. The comparator detects when the voltage of the capacitor matrix has been discharged below a reference voltage, and then produces the output signal after the desired delay, as illustrated in Figures 1 and 2.
Figure 1. Digital-to-time converter (DTC) for TDC.
According to the simulations, a linear dynamic range of 12-13 bits, i.e. several ns, with sub-ps resolution can be achieved for the DTC. This means that an interpolator with a linear dynamic range of 5 ns can be combined with a 200 MHz counter to implement an integrated CMOS TDC with sub-ps level resolution and a ms-level dynamic range.
Figure 2. Operation of DTC.
A CMOS multi-channel time-to-digital converter (TDC)
A 7-channel time-to-digital converter (TDC) that has been developed and realized is able to measure the time intervals and pulse widths or rise times of several successive timing signals, Figure 3 [2,6]. A multi-channel measurement structure is required in pulsed laser TOF measurement when several pulse echoes arrive at the receiver, and in order to compensate the timing walk error which explained below.
Time digitizing is based on a counter and a two-level stabilized delay line interpolation. The 14-bit counter counts the full reference clock cycles between the timing signals, which makes a long microsecond-level measurement range possible. The interpolators find the locations of the timing signals within the reference clock cycles with much higher resolution. The effects of process, voltage and temperature variations are cancelled with continuous automatic delay-locked-loop (DLL) based stabilization methods. The time interval measurement unit is fully integrated and only a low frequency crystal is needed as an external component.
The performance of the circuit represents the state of the art with regard to precision among wide-range TDCs. The TDC offers a precision better than 10 ps for all the required measurements (time intervals and pulse widths) and the measurement range is up to ±74 µs, see Figure 4. The performance is based on optimizing the three main factors, which affect the precision. The main factor usually affecting negatively the precision, interpolation nonlinearity, was minimized with an interpolation structure based on the reference recycling method. The recycling technique uses a very short delay line for the interpolation, which reduces the accumulation of delay errors in the interpolation. The quantization error was minimized with the second interpolation level, which uses parallel capacitor-scaled delay line structures in order to digitize time intervals with higher than 10 ps resolution. The high speed (flash-type) measurement method minimizes the injurious effect of jitter on the measurement signals because the measurement result is ready immediately after the arrival of the timing signals, and hence the jitter sources do not have time to cause error.
Figure 3. Block diagram of the 7-channel CMOS TDC.
Figure 4. Measurement example with start and 3 stop pulses.
Oscillator instability effects in time interval measurement
The impact of phase noise on time interval measurements has been studied. Time-to-digital converters utilize a reference clock to provide a stable time base for time interval measurements. Today’s TDCs aim at a precision close to one picosecond, which sets very high requirements on the reference time base.
Phase noise, and thus jitter, are non-stationary processes, which are usually modeled in the frequency domain using power-law functions. Time domain modeling is slightly more difficult, so the jitter needs to be differentiated once or twice to make the noise process stationary. Other techniques take into account that the jitter is observed only for a finite duration. Furthermore, flicker noise and white noise require that one knows the noise bandwidth.
Using these techniques, time domain representation of time interval jitter for various noise processes was introduced; this can be used to evaluate the suitability of an oscillator in time interval measurement. The time domain statistics can be directly calculated from phase noise measurements. Also a simple measurement technique was introduced which allows us to estimate the total white noise by converting the clock signal into a square wave. Ideally, this conversion rejects AM, and PM is sampled at the clock edges, which also makes the noise process cyclostationary. Because of this, all the original PM power is folded into lower frequencies, which makes it possible to easily estimate the jitter due to white noise and flicker noise. Flicker noise also usually modulates the frequency of an oscillator. It was shown that the time interval jitter due to this kind of noise is proportional to the time interval and the observation time.
Spurious tones were also shown to have a large impact on measurement precision, which sets tight constraints for a PLL or DLL if one is used in synthesizing the reference clock. Spurious tones introduce a time interval dependent error, which can be erroneously interpreted as a non-linearity error in a TDC. All these results were confirmed by measurements done with a TDC capable of measuring up to 1ms with a precision of 2 ps .
A laser scanner chip set for accurate perception systems
Two integrated circuits were developed that realize the receiver channel and time interval measurement functionalities of a pulsed time-of-flight laser radar. The circuits, a receiver channel circuit (REC) and a time-to-digital converter circuit (TDC, above), were used in a novel miniaturized laser scanner sensor shown in Figure 5 to serve numerous applications especially in vehicular perception systems (www.minifaros.eu) .
Traffic perception is a demanding application for several reasons. Several laser scanning functionalities from pedestrian detection to adaptive cruise control require cm-level precision in a wide measurement range. The measurement performance needs to be maintained, also with the highly varying reflectance properties of the target (> 1 : 10 000). Varying laser echo amplitude creates error in the time detection moment because the detection threshold is constant. This detection error, also known as walk error, is usually the largest error source in laser distance measurement. A problem arises also when several pulses arrive at the receiver from the cover lens, because of fog or rain for example, before the actual echo from the measurement object.
The REC-TDC circuit pair forms the digital measurement result equivalent to the distance between the LIDAR sensor and the measurement target as follows. The transimpedance amplifier in the REC circuit receives the photocurrent pulse from the APD and generates an equivalent output voltage pulse. A timing comparator compares the signal with a predefined adjustable reference voltage level. As soon as the signal crosses the reference level, the comparator generates logic level stop signals for the TDC. Both the rising and the falling edges of the received pulse are detected.
The TDC measures the time intervals between electrical timing pulses by using counter and delay line-based interpolation techniques and converts the results to digital words. The TDC has 7 measurement channels, and thus can measure the time intervals from a start timing signal to 3 succeeding stop signals, which correspond to the distances from the laser scanner to the target. In addition, it measures also the pulse widths of the stop signals. The pulse width has a relation to the input amplitude-dependent timing error (walk-error), which means that the stop signal pulse width information can be used in the timing walk error compensation. The power consumption of the REC and TDC circuits are 130 mW and 85 mW, respectively. The block diagram of the REC circuit is shown in Figure 6 .
Figure 5. A LIDAR sensor, APD receiver PCB and the developed receiver channel (REC) and time-to-digital converter (TDC) circuits. The REC IC (1.6 mm x 1.6 mm) was fabricated in a 0.35 μm SiGe BiCMOS technology and the TDC IC (2.4 mm x 3.7 mm) was fabricated in a 0.35 μm CMOS technology. The circuits were packaged in plastic QFN32 and QFN36 packages, respectively.
Figure 6. Block diagram of the receiver channel.
The transimpedance and the bandwidth of the receiver channel were measured to 64 kΩ and 300 MHz, respectively. The REC circuit input RMS noise was measured to be about 100 nARMS. The timing walk error measurements were performed on a test bench where the amplitude of the optical pulse could be varied over the range of 1 : 22 000 with a neutral density filter. The compensation curve, realized as a lookup table (LUT) and shown in Figure 7a, shows the dependence between the generated walk error and the measured pulse width. The pulse width widens monotonously throughout the dynamic range measured. The walk error without compensation was about 2.2 ns, which corresponds to about 32 cm in distance. The walk measurements were performed for different start-stop delays, and the walk was compensated for by means of the LUT. Compensated walk errors, shown in Figure 7b, were less than ±20 ps (corresponding to ±3 mm in distance) over the amplitude range of 1 : 22 000.
Figure 7. a) Measured compensation curve over the dynamic range of 1:22 000, b) Residual walk errors for different start-stop time delays, c) Single-shot precision.
The single-shot precision was determined by recording 2000 single-shot measurements and calculating the standard deviation of the timing point of the rising edge. In addition, each individual measured result was compensated for by means of the walk error compensation curve and the single-shot precision was calculated from the obtained distribution. The worst case single-shot precision is shown in Figure 7c as a function of input amplitude. This result includes the jitter of the TDC, the jitter from the receiver, the jitter introduced by compensation and the jitter caused by the measurement environment . The worst case single-shot precision was about 144 ps (±10 mm). At high input signal levels, the single-shot precision was about 14 ps. As can be seen from Figure 7c, the single-shot precision after compensation is better than the jitter measured from the rising edge. The reason for this is that the walk error compensation principle based on the pulse width measurement compensates also for the timing jitter in the laser pulse caused by the laser pulse amplitude jitter (laser shot noise).
Multiphase time-gated 2x128 single photon avalanche diode (SPAD) array for Raman spectroscopy
Raman spectroscopy is based on inelastic scattering, or Raman scattering, of monochromatic light, usually from a CW (continuous wave) laser in the visible, near infrared or near ultraviolet range. Unfortunately, the Raman spectrum is masked in most otherwise potential cases by a strong fluorescence background. This is due to the fact that the probability of Raman scattering is much lower than that of fluorescence. As a result, in spite of the obvious advantages of Raman spectroscopy, this strong fluorescence background has so far restricted its use in most potential applications in the fields of the agricultural, food and oil industries, security control and crime investigations, for example.
It seems to be possible to suppress the fluorescence background to a great extent if intensive short laser pulses are used to illuminate the sample instead of CW radiation, and by recording the sample response only during these short pulses. The suppression is due to the fact that Raman scattering is introduced immediately after the collision between the photons and the sample material, unlike fluorescence, which is emitted after a delay characteristic to the sample. Thus, by “time-gating” the measurement for only the period of the laser pulse, most of the fluorescence is blocked out from the recorded spectrum .
The block diagram of the pulsed Raman spectrometer and the principle of fluorescence suppression are shown in Figure 8. The material to be measured is excited by means of a pulsed laser, emitting short 70-150 ps (FWHM) and spectrally narrow (~0.2 nm) pulses at an average power level of a few mW. The SPADs of the detector array (SPAD-IC) are enabled by the trigger pulse from the pulsed laser just before each laser pulse, and the photons are counted during short time periods (Dt1, Dt2, Dt3, Dt4), in order to suppress the fluorescence and effective dark count rate (DCR). In addition to fluorescence and DCR suppression, the time constant of fluorescence can also be determined.
Figure 8. Block diagram and time gating principle of the proposed Raman spectrometer.
A Multiphase time-gated 2*4*128 single photon avalanche diode (SPAD) array has been designed and fabricated with standard high voltage 0.35 µm CMOS technology for Raman spectroscopy. Each of these two columns consists of four parallel connected SPADs. A 50 ps -100 ps time resolving capability can be achieved by using this multiphase time-gated structure. The multiphase time-gated structure makes it possible to suppress the fluorescence background by a factor of 10 -100, depending on the time constant of fluorescence and the value of the time-gate used. The dark count rate of the SPAD is also minimized by means of the time-gating in order to maximize the signal-to-noise ratio of the Raman signal. The size of the SPAD matrix and the whole IC are 4100 μm x 343 μm and 4300 μm x 2950 μm, respectively. The fill factor of the designed SPAD matrix is 23%. A photograph of the SPAD array IC is shown in Figure 9.
Figure 9. Photograph of multiphase time-gated 2x128 SPAD array.
Generation and detection of electrical/optical transients
High-energy picosecond laser diode pulse generation by “enhanced gain switching”
A previously proposed strongly asymmetric DH bulk laser diode, with an extremely large “equivalent spot size” served as the fundamental structure for further experimental investigations. The implementation of a saturable absorber area into the laser diode cavity by focused ion beam (FIB) technique (Figure 10) affects positively the optical pulse shape, outputting in “enhanced gain switching” operation mode up to 50 W, 61 ps laser pulses (Figure 11), measured with a 25-GHz pin photodetector. The structure utilizes a relatively simple driving scheme provided by an avalanche transistor current pulser (Figure 12) intended for laser ranging and other optoelectronic measurement applications, especially for single photon measurements.
Figure 10. SEM graph of the laser diodes cavity front before and after saturable absorber area implementation; (cavity length: 1.4 mm, oxide stripe width: 124 µm)
Figure 11. Measured driving current and corresponding optical output pulses of a DH bulk high-speed laser diode based on “enhanced gain-switching” (app. 850 nm).
Figure 12. Laser diode and driver board (area app. 1 cm2).
Sub-mm precision time-of-flight distance measurement with a gain-switched semiconductor laser and single-photon avalanche diode (SPAD) detector
A sub-millimeter precision distance measurement capability of the single photon time-of-flight measurement method was examined. The above new type of gain-switched laser diode and a SPAD detector were used, with a distance measurement target attached to a stepper motor alternating between positions 0.0 mm and 2.0 mm. A moving target was chosen to eliminate the long term measurement drift. The measurement result in Figure 13 shows that a ~50 µmrms distance measurement precision can be achieved in a measurement time of 1 second at a measurement rate of 105 signal counts per second. Compact laser pulsers with fast pulsing rates that enable these types of measurements are also being worked on .
Figure 13. 38 distance measurement results as a function of the amount of averaged single-photon flight time results, and their standard deviation.
High-speed avalanching BJTs as high-current drivers for LDs and UV LEDs, and as high-power pulsed emitters for sub-THz imaging
Currently, a principle problem in developing a unique (high-voltage, picosecond range) GaAs avalanche switch lies in premature surface breakdown, which does not allow the device to be biased close to the breakdown voltage in the bulk. At first glance, this limitation is of a fundamental character, as the premature surface breakdown in the n+-p-n0-n+ GaAs avalanching BJT (ABJT), with n+ emitter on the top, is unavoidable due to the negatively-beveled mesa contour of the p-base-n0 collector junction. We have, however, found a new mechanism stabilizing the surface current on a reasonably low level due to intrinsic surface traps, thus permitting biasing close to the bulk breakdown voltage to be used. This extremely favorable and important phenomenon, termed “avalanche-assisted carrier trapping”, consists of electron trapping on the deep surface acceptors, whose negative charge broadens the space-charge region on the surface, thus reducing the peak of the electric field and preventing sharp growth in the surface current . In addition, we have lately suggested a new concept for passivation of a mesa-surface using a specially formed negative surface charge. We realized this idea in an experiment using simple deposition of massive chalcogenide glass on the mesa-surface of a GaAs ABJT and gave physical interpretation for the passivation mechanism, see Figure 14 .
For nanosecond/sub-nanosecond pulse durations, GaAs-based ABJTs are apparently the best active high-voltage/high-speed switches, but their technology has been thus far under development, and commercialization of this device still requires significant investments and time. Si ABJT’s have been most frequently used for nanosecond pumping of pulsed laser diodes, but Si avalanche transistors with a switching time of less than 2 ns are absent. Highly surprising is the fact that a comprehensive physical description of Si ABJT operation at high current densities was absent until the last decade, and the first reliable results we have obtained a number of years ago using 1-D and 2-D approaches. Very recently, we have also found the drastic effect of 3-D phenomena [12, 13] on the device operation, especially in short-pulsing mode. Several interesting findings were made concerning significant and non-monotonic change in the switching size along the emitter-base fingers during the transient due to two competing physical mechanisms: channel shrinkage and turn-on spread (see Figure 15). Principally important is the peak current density value, which has to be as high as possible for fast switching with low residual voltage, but at the same time should not exceed the destruction level due to local overheating. Obviously the drastic effect of 3-D dynamics is of principle importance and cannot be ignored when designing unique high-speed Si avalanche transistors, which could generate the pulses in a sub-nanosecond range at an amplitude of 1-10 A.
Together with high-current/short pulse generation, a very promising (and apparently highly important) application for the avalanche switching in GaAs BJT is the generation of pulsed broad-band terahertz emission. Periodical nucleation and annihilation of ultra-narrow, powerfully ionizing “collapsing” domains is believed to cause the THz emission observed in our experiments. The task of design, development and investigation of high-power pulsed (ns/sub-ns) emitters for a new generation of active sub-THz imagers is very complicated and challenging, and should be divided into several directions and stages. Very lately we have undertaken an attempt to realize “collapsing” field domains (causing sub-THz radiation) not in a GaAs bipolar transistor structure, but in an analogue of a Gunn diode with special impurity profiles, and under special pumping conditions. Modeling results are promising, and we plan to perform their experimental verification. The main direction, however, is design and development of BJT GaAs-based structures, combined with properly designed sub-THz antennas, using a room-temperature quasi-optical detector based on a Schottky diode. The solution of a large number of the related tasks is underway, and the first laboratory examples of transmission sub-THz imaging utilizing not only transmission intensity, but also propagation delay of the pulses across the object with temporal resolution of ~10-30 ps have been presented in several plenary and invited talks.
Figure 14. Illustrations of (a) premature surface breakdown caused by narrowing of the space-charge region (SCR) on the mesa surface intrinsic to a negative beveled angle (φ‑); (b) suppression of surface breakdown by a negative charge Nsc on the mesa surface due to SCR broadening; (c) “soft” surface breakdown caused by the trapping of impact-generated electrons on the surface. The latter mentioned mechanism sustains surface breakdown over a broad collector voltage range, though limiting the surface current to a moderate (non-destructive) level.
Figure 15. Switching transient (a), current density profiles (b) and normalized current density profiles (c) simulated for an avalanche transistor. Instants corresponding to the curves in (b) and (c) are related to the transient in (a) as follows: 1-4.2 ns; 2-6.2 ns; 3-6.8 ns; 4-7.3 ns; 5-8.4 ns; 6-9.3 ns. Curve 7 in (c) shows the shape of the optical excitation. Please note that normalized curves in (b) first shrink from the initial excitation profile (curve 7) to the current density profiles 1 and 2, and then spread to profiles 3, 4, 5 and 6.
Circuit analysis and linearization techniques
Distortion contribution analysis algorithm VoHB
A general-purpose distortion contribution analysis called VoHB (Volterra on Harmonic Balance) is being developed, and in 2009 it appeared as a functional, multi-device, model-independent prototype within an AWR-Aplac’s analog simulator.
VoHB runs at VCCS (voltage-controlled voltage source) level, and to be able to calculate frequency-domain mixing effects, each VCCS is modeled by a polynomial that is fitted using the simulated large-signal spectra obtained directly from harmonic balance simulation. Some of the numerical fitting techniques were reported at the SCEE conference (Scientific Computing in Electrical Engineering) in Zurich, September 2012, and a previously conducted comparison of time-varying and time-invariant Volterra analyses was published .
It was recently found that when new device models are distributed as executable Verilog-A code, their actual VCCS level implementation depends on the software performing this conversion – there may be intermediate nodes or additional loops to solve some intermediate value, for example. This makes the contributions more complex to interpret, and in 2012 the first efforts were made to reduce the distortion contribution to transistor input and output ports, in a similar fashion as input reduced noise is described. The first results were reported at the SCEE 2012 conference.
Design methods for switching amplifiers
Switch-mode RF amplifiers have been studied, especially the design method of class E amplifiers with non-linear output and feedback capacitance . The experiences so far were summarized in Simo Hietakangas’s doctoral thesis, which was successfully defended in June 2012. The main findings were related to strongly varying input impedance of supply-modulated amplifiers in general, and especially switching amplifiers. Switching amplifiers were also shown to inject strong 2nd harmonic to their input node, making a very low impedance drive necessary to keep the duty cycle constant.
The research has continued by further development of the test setup of supply modulated amplifiers. A new supply modulator voltage source has been developed, and a great deal of effort has been spent on recognizing and minimizing the effects of output ripple and spurs.
Biomedical and power management
Co-operation on measuring the nerve signals of living insects continued with the Dept. of Physics. During 2012, a PLoS One paper was published . It describes a micro-positioning system that actively cancels the movements of the insect, hence keeping the probe exactly in the intended nerve cells.
Recently, work in the field of energy harvesting applications has also been started by one doctoral student. Christian Schuss wrote three conference papers this year, mainly concentrating on the shadowing effects on car roof-top photovoltaic (PV) panels and the possibility to bypass shadowed panels. Also an extensive measurements set of outdoor illumination and PV I-V responses was collected during the summer.
Digital error correction on A/D and D/A converters
In February Marko Neitola defended his doctoral thesis on spurious responses in delta-sigma modulators. The main results were related to an easily usable generalization of the data weighted averaging (DWA) algorithm, and to the origin of mismatch spurs in multi-bit converters when using DWA technique. It was shown that most of the spurs can be moved out-of-band simply by re-ordering the unit DAC elements so that even-order harmonics are minimized in the Fourier transform of the DNL of the unit DAC. Figure 16 illustrates the location of the spurious tones vs. input voltage amplitude.
One of our doctoral students, Jia Sun, has developed a model where continuous-time settling effects in a switched-capacitor circuit can be combined in discrete-time simulations of, for example, a delta-sigma modulator. Previous studies have assumed isolated stages and certain clock phasing; the new method can handle also cascaded switched-capacitor stages that settle simultaneously. The technique is based on pre-characterizing the settling error with a limited set of initial conditions, and it was reported in detail at the Norchip 2012 conference .
Figure 16. Frequency of spurious tones vs. input amplitude due to DNL non-linearity in a DWA-linearized delta-sigma modulator (DSM). a),b) a low-pass DSM, c),d) a band-pass SDM. Both DC sweep and swept-amplitude sinusoidal input used.
Academy of Finland
Ministry of Education and Culture
other domestic public
Alahdab S, Mantyniemi A & Kostamovaara J (2012) A 12-bit digital-to-time converter (DTC) with sub-ps-level resolution using current DAC and differential switch for time-to-digital converter (TDC). Instrumentation and Measurement Technology Conference (I2MTC), 2012 IEEE International, 2668–2671. 
Jansson J, Koskinen V, Mäntyniemi A & Kostamovaara J (2012) A multi-channel high precision CMOS time-to-digital converter for laserscanner based perception systems. IEEE Transactions on Instrumentation and Measurement 81(9): 2581-2590. 
Keränen P & Kostamovaara J (2012) Oscillator instability effects in time interval measurement. accepted to be published in IEEE Transactions on Circuits and Systems (TCAS I). 
Kostamovaara J, Kurtti S & Jansson J-P (2012) A receiver - TDC chip set for accurate pulsed time-of-flight laser ranging. Proceedings of the CDNLive!EMEA2012 conference, May 14-16, Muenchen, Germany, 4 p. 
Kurtti S, Jansson J-P & Kostamovaara J (2012) A laser scanner chip set for accurate perception systems. Proceedings of the 16th International Forum on Advanced Microsystems for Automotive Applications-AMAA 2012, May 30-31, Berlin, Germany, 313-322. 
Jansson J-P, Koskinen V, Mäntyniemi A & Kostamovaara J (2012) A multi-channel wide range time-to-digital converter with better than 9ps RMS precision for pulsed time-of-flight laser rangefinding. Proceedings of the ESSDERC/ESSCIRC 2012, September 12-16, Bordeaux, France, 4 p. 
Nissinen J, Nissinen I & Kostamovaara J (2012) An integrated CMOS receiver-TDC chip for mm-accurate pulsed time-of-flight laser radar measurements. IEEE International Instrumentation and Measurement Technology Conference (I2MTC), 421-424. 
Lanz B, Kostamovaara J, Vainshtein S, Lantratov V & Kalyuzhnyy N (2012) Single-heterostructure laser diode producing a 6W/40ps optical pulse from a 20μm stripe width. Opt. Eng. 51(5): 050503(1-3). 
Lanz B, Vainshtein SN, Lantratov VM, Kalyuzhnyy NA, Mintairov SA & Kostamovaara JT (2013) Picosecond internal Q-switching mode correlates with laser diode breakdown voltage. Semiconductors 47(3): 383-385, Springer. 
Hallman LW, Haring K, Toikkanen L, Leinonen T, Ryvkin BS, & Kostamovaara JT (2012) 3 nJ, 100 ps laser pulses generated with an asymmetric waveguide laser diode for a SPAD TOF range finder application. Meas. Sci. Technol. 23: 025202, 8. 
Vainshtein S, Javadyan V, Duan G, Tsendin K, Hovhannisyan R & Kostamovaara J (2012) Chalcogenide glass surface passivation of a GaAs bipolar transistor for unique avalanche terahertz emitters and picosecond switches. Applied Physics Letters 100(7): 073505-073505-4. 
Duan G, Vainshtein S & Kostamovaara J (2012) Turn-on spread determines the size of the switching region in an avalanche transistor. Applied Physics Letters 100(19): 193505-193505-4. 
Duan G, Vainshtein S & Kostamovaara J (2012) Three-dimensional peculiarities in an avalanche transistor provide a broadened range of amplitudes and durations of the generated pulses. Applied Physics Letters 101(17): 173506-173506-4. 
Rahkonen T & Aikio J (2012) Comparison of ordinary and time-varying Volterra analysis for finding distortion contributions. Springer j. Analog Integrated Circuits and Signal Processing. 
Hietakangas S & Rahkonen T (2012) Analysis of class E amplifier with both linear and nonlinear shunt capacitance at any duty cycle, grading coefficient and supply modulation. Published electrically in Springer j. Analog Integrated Circuits and Signal Processing 71(2): 245-253. 
Kursu O, Tuukkanen T, Rahkonen T & Vähäsöyrinki M (2012) 3D active stabilization system with sub-micrometer resolution. PLoS ONE: e42733. 
Sun J, Neitola M & Rahkonen T (2012) Behavioral modeling of nonlinear settling for multiple cascaded SC stages. Norchip 2012 conference, November, Copenhagen, Denmark. 
Last updated: 22.6.2016