Coded data generation or conversion – Analog to or from digital conversion – Analog to digital conversion
Reexamination Certificate
1998-12-14
2001-03-06
JeanPierre, Peguy (Department: 2819)
Coded data generation or conversion
Analog to or from digital conversion
Analog to digital conversion
C341S159000
Reexamination Certificate
active
06198420
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to analogue tracking amplifiers, especially such as peak/valley detectors used in digital data receivers such as fiberoptic receivers. It also relates to flash analog to digital converters.
In a fiberoptic network, information is transferred as pulses of light across fiberoptic cable links. Optical networks typically include transfer points called “mixers” or “couplers” which accept signals from one fiberoptic link and transfer that signal to another link. Because practically all computers and data processing devices use electrical signals to operate, most origin and destination points in fiberoptic networks include converters that either detect optical signals and then produce electrical pulses, or detect electrical pulses and generate corresponding optical signals on the optical network. Converters located at destination points usually consist of a photodetector, a preamplifier and a data detection circuit.
Fiberoptic receivers typically comprise an optical preamplifier which takes the incident optical data and converts it to a voltage waveform whose amplitude is proportional to the intensity of the light pulses. The intensity of the light pulses tends to be poorly defined as it depends on the location and optical performance of the transmitter or transmitters. Typically multiple transmitters, each of which will transmit during different time intervals, are present. Furthermore, the voltage level at the output of the optical preamplifier, which corresponds to a dark signal or absence of an incident light pulse, is often very poorly defined and varies widely from preamplifier to preamplifier. It, however, tends to be relatively stable for any one optical preamplifier, except for temperature variations and component aging effects. The characteristics of the output signal of an optical preamplifier will have relatively time invariant dark (or signal zero) levels and varying amplitude pulses (corresponding to different active transmitters) which are superimposed on this dark level.
A second element of a fiberoptic receiver system takes the analog output of the preamplifier and recovers both clock and data information from the analog waveform. The first element of this clock and data recovery system is a comparator whose trigger level is set at approximately fifty percent (50%) ofthe amplitude of the signal. The comparator slices the signal into logic ones and logic zeros. The comparator evaluates the signal at a sample rate equal to or at some multiple of the data rate. When a sample rate of a multiple of the data rate is used, the signal is said to be oversampled and, in this case, it is not necessary for the clock to have a known phase relationship with the incoming data.
Consequently, because of the unknown dark level and the characteristics of the output signal, the data from the optical preamplifier has some characteristics that make subsequent processing of the signal difficult.
FIG. 1
illustrates a waveform
11
of a received stream of logic zeros
13
and logic ones
15
with five different signal amplitudes, TX
1
, TX
2
, TX
3
, TX
4
and TX
5
all sharing a common dark level
17
. The different signal amplitudes correspond to five different transmitters on a fiberoptic network. The switch over from one transmitter to another is usually accomplished within one bit period. This means that the receiver circuitry must quickly sense the change and respond to a second output signal such as a TX
2
having a logic one represented by a signal with a different amplitude from the previous signal, TX
1
.
A second difficulty is that the DC average value of the signal varies with each transmitter and so it is impossible to use AC or capacitor coupling of the signal to establish an average value that can be used with a comparator to decide which level represents a logic zero and which level represents a logic one.
FIG. 2
illustrate the above point showing the effect of AC coupling on a multi-amplitude data stream.
FIG. 2A
includes a signal waveform
27
from the optical preamplifier and shows peak
23
which corresponds to the logic ones while valley
21
corresponds to logic zeros.
FIG. 2B
illustrates the effects of capacitance coupling. The waveform
27
now floats on a DC average signal represented by dash line
25
and consequently the peak
23
and valley
21
no longer correspond to logic zeros or logic ones (reference should be made to point
28
). It should be noted that when the data amplitude changes, the average of the data pulses is lost for a period related to the time constant of the AC coupling system.
The presence of multiple amplitude signals in the data requires the use of multiple comparators or an analog to digital converter whose bit weighting is sufficiently small so as to adequately slice the smallest amplitude signal and whose full scale voltage is sufficiently large so as to adequately slice the largest signal amplitude at approximately their respective fifty percent (50%) values. In practice, the above requirements place severe restrictions on the analog to digital converter's resolution and speed. For example, in a 5V system, the dark level signal from a different optical preamplifier varies from 0.5V to 4.0V and the signal amplitude varies from 50 mV to 0.5V. The bit rate is typically in excess of 100 Mhz. An analog to digital converter that can cover this range would need a bit weighting of 25 mV and a measurement range from 0.5V to 4.25V implying a resolution in excess of 6 bits at a minimum of 100 Mhz conversion rate.
In U.S. Pat. No. 4,431,916 to Couch, Couch uses a signal delaying circuit to create threshold signals and detects the peak of a first delayed signal between the one-half rise point in the input signal and a one-half fall point in a second delayed signal.
Devices exist for detecting a predetermined voltage level and for using the detected voltage level to offset a circuit. U.S. Pat. No. 3,736,582 to Norris discloses a compensation circuit that adjusts a baseline voltage up and down in a system where that baseline voltage “gallops” or changes.
U.S. Pat. No. 5,381,052 issued to Kolte discusses a device which detects and tracks a peak level through a network of operational amplifiers and a capacitor which holds the peak voltage value. Kolte's device inverts a detected input voltage and drives the capacitor in two different states, depending on the polarity of the voltage between input and the inverter. In a first state, the capacitor follows the input voltage, while in a second state, the capacitor value is the average of the input voltage and the inverter voltage.
SUMMARY OF THE INVENTION
The present invention addresses the above problems by providing a dark level regeneration circuit that is fully integrated into an analog to digital converter front end. The analog to digital converter zero level, or reference level, is set to the dark level and hence the effective conversion range of the analog to digital converter is restricted to half of the maximum expected signal amplitude. In the above example, the resolution is now reduced to between 2 and 3 bits, thus, substantially reducing the complexity, power consumption and cost of the analog to digital converter. A further advantage of the invention is that the reference voltage of the analog to digital converter can be referenced to ground (or any other convenient voltage) and does not have to track the DC component of the signal. This advantage provides additional circuit simplification.
The present invention provides an analog to digital signal converter which has a signal input, a reference potential input, and a dark detector for providing a normalized reference signal. The digital signal converter possesses a threshold voltage generator connected to the reference potential input which produces at least one threshold voltage.
The analog to digital signal converter also has a system of preamps and regenerative comparators which compare a signal with respect to its dark level to a threshold voltage which is defined a
Keane John A.
Ryan John G.
van Ettinger Rudolf G.
Jean-Pierre Peguy
Robinson Richard K.
Silicon Systems Research Limited
LandOfFree
Multiple level quantizer does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Multiple level quantizer, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multiple level quantizer will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2460050