Driverless doesn't just reduce chip-count, it also reduces power dissipation.
Here's a quote from a section in the "Advantages of Lightwave Logic Polymer Technology for Photonics Applications" Photonics Guy wrote it.
Quote:
D. Reduced Drive Voltage Requirement (Low Vpi)
LWLG's polymers provide a significantly lower modulator voltage for biasing device (Vpi). Low Vpi operation is a critically important parameter as there is a direct correlation to the amount of power required to operate the devices. Traditional LiNbO3 modulators have Vpi's up to 10V. The IP LWLG has recently been granted enable Vpi's as low as 0.5V. The lower Vpi allows the modulator to be driven directly from CMOS circuits and does not require amplification to boost the signal levels to higher voltages. The elimination of the amplifier reduces power, increases signal integrity and - consequently - provides increased optical bandwidth and faster data rates as well as price reductions.
Here is a good App Note from ixBlue that describes what a Modulator Driver is in fairly good detail.
The modulator Driver does two things, (1) it amplifies the incoming RF signal (the datastream) to the level that is optimized for the modulator (dependent upon the Vpi of the modulator). (2) it tracks the optimum DC bias point of the modulator (modulation dependent) and maintains the bias by tracking the output in a feedback loop to address operating point drift w.r.t. temperature, signal levels, electrode capacitance, etc. So the driver needs to have a very linear high bandwidth RF amplifier, DACs/ADCs, op-amps and a processor to do all these fine functions.
The most power-hungry component in the Driver will be the RF Amp, typically using GaAs transistors. Linear amplifiers are notoriously inefficient (Class A). And - to be linear - need to be 'backed off' from their 1dB compression point (P1dB) by several dB. It is difficult (read expensive) to build a wideband RF amplifier that has linear characteristics, a flat passband and low noise. The difficulty in building a high-performance RF Amplifier is compounded as the bandwidth is increased. With LWLG's modulators approaching 100 GHz, this becomes an extremely challenging (expensive) endeavor.
If the Vpi of the modulator is low enough, the incoming signal does not require amplification, which obviates the need for this expensive, power-hungry Driver component. Most Modulator Drivers require 12V to operate and consume prodigious amounts of power. Removing the need for the driver saves a LOT of power and expense. This is true regardless of whether the modulator is intensity or phase modulated (amplitude or phase modulation). Consequently, Direct Drive (Driverless) Modulators such as LWLG is marketing are a really, really big deal to the industry.
Not only does the low Vpi of 0.5-1.0V allow a modulator to be driven directly from CMOS circuits, it eliminates the requirement for the use of high voltage, inefficient voltage regulators and other biasing components. Power dissipation goes up proportionally to the square of the voltage used. If you have to use 5V to power your circuit, that circuit can consume up to 25X the power of a circuit that only runs on 1V. Humongous power savings!!!
PG
This post was compliments of Photonics Guy
X
