Phase-compensated impedance converter

Details Phase-compensated impedance converter Phase-compensated impedance converter

1999-12-23

2002-06-11

Lee, Benny (Department: 2817)

Wave transmission lines and networks

Negative resistance or reactance networks of the active type

Simulating specific type of reactance

C331S10800D

Reexamination Certificate

active

06404308

ABSTRACT:

BACKGROUND
This invention relates to electronic impedance converter circuits and more particularly to impedance invertors, or gyrators, and even more particularly to phase-compensated gyrators and integrators.
One of the advantages of electronic integrated circuits (ICs) is small size, and so ICs have become ubiquitous in hand-held and other equipment, but some circuits require components that are difficult to integrate. In particular, it is difficult to integrate passive inductors having impedances of more than a few nanohenries (nH). Thus for example, on-chip passive signal filters are normally limited to resistor-capacitor (RC) filters, except for filters designed for very high frequencies where inductors (coils) of a few nH are effective.
In communications and electronics, a filter is generally hardware or software that selectively passes certain elements of a signal and eliminates or minimizes others. A filter in a telecommunication network, for example, may transmit signal elements either up to a certain frequency and attenuate (dampen) those above it (a low-pass filter), or down to a certain frequency and attenuate those below it (a high-pass filter), or within a band of frequencies (a band-pass filter).
It is possible to overcome the limited impedances of integrable passive inductors by using combinations of active devices (e.g., op-amps), resistors, and capacitors that are easily integrable. One combination that can mimic the properties of a passive inductor is the integrator, which can generally be considered to convert an impedance to its inverse. Integrators are often used in discrete-time (digital) filters. Continuous-time (analog) filters implemented with integrators typically employ such elements in loops and these loops are often connected back-to-back. Two integrators in a loop actually form a gyrator, and if the forward and backward integrators have the same transfer characteristics they form a passive gyrator, otherwise they form an active (or asymmetric) gyrator.
Gyrators are described in the literature, which includes P. Horowitz et al.,
The Art of Electronics
2d ed., pp. 266-267, 281 Cambridge University Press (1989); Fink et al.,
Electronic Engineer's Handbook
4th ed. (D. Christiansen et al. eds.), pp. 16.38-16.39 McGraw-Hill, New York (1997); and B. Nauta, “A CMOS Transconductance-C Filter Technique for Very High Frequencies”,
IEEE J. Solid-State Circuits
, vol. 27, no.2, pp. 142-153 (February 1992). Various circuits employing gyrators and integrators are described in German Patent Application No. 199 364 30-3 filed on Aug. 3, 1999, by S. Mattisson for “Analog Filter”; and U.S. patent application Ser. No. 09/274,327 filed on Mar. 23, 1999, by S. Mattisson for “Demodulator Circuits”, both of which are expressly incorporated in this application by reference.
In general, the output signal produced by a gyrator is delayed in time and is shifted in phase with respect to the input signal provided to the gyrator. As seen from
FIGS. 1A
, B, a gyrator generally comprises a positive transconductance g
m
and a negative transconductance −g
m
that receive input voltage signals v
1
, v
2
and produce a combined output current signal.
Gyrators can be implemented in various circuits. For example, a gyrator can be implemented in a combination of inverter circuits as illustrated in
FIG. 2
, which depicts a gyrator core
10
having four invertors
12
,
14
,
16
,
18
arranged as described in the above-cited Nauta publication. The negative transconductance is realized by employing differential input signals i_
1
, i_
2
and differential output signals o_
1
, o_
2
and by crossing one pair of output-to-input connections. This crossing produces a loop through the four invertors (transconductances), and it will be appreciated that although
FIG. 2
shows the connection between inverter
12
's output and inverter
16
's input crossing the connection between inverter
18
's output and inverter
14
's input, it is possible for the loop to be formed by crossing the connection between the inverter
12
's input and the inverter
16
's output and the connection between the inverter
14
's output and the inverter
18
's input.
Such a filter typically may also include sections
20
,
30
for feeding back common-mode voltages. As depicted in
FIG. 2
, section
20
has four invertors
22
,
24
,
26
,
28
and section
30
has four invertors
32
,
34
,
36
,
38
.
Crossing connections as depicted in
FIG. 2
can cause stability problems in circuits using such gyrator cores. It will be appreciated that a stability analysis of a gyrator-based filter such as that depicted in
FIG. 2
is substantially the same as the stability analysis of an integrator-based filter because integrators are parts of gyrator loops. First-order stability analysis such as that described in the Nauta publication reveals that stable, i.e., non-regenerative, behavior can be obtained only if invertors are used in the gyrator core
10
. For example, the transconductances, e.g., invertors
12
,
14
,
16
,
18
, may be single metal-oxide-semiconductor (MOS) transistors, including complementary MOS (CMOS) field-effect transistors (FETs). Using MOS components, and in particular CMOS components, has a number of advantages, not the least of which is low power consumption.
More detailed analysis of the MOS transistors, however, shows that non-quasi-static behavior of channel charges in each MOS device adds a delay that can be approximated by a parasitic pole in the frequency response of the transconductance of each device. This extra pole, or delay, makes the gyrator unstable if it is not designed properly. The actual performance of the simple filters described in the Nauta publication generally accords with Nauta's predictions, but the actual performance of the complicated filters described by Nauta may deviate by as much as 10 dB from the predicted behaviors.
Nauta's filters could be stabilized by separate Q-tuning circuits (e.g., a separate supply voltage for ballast invertors in common-mode feedback networks) that could be externally adjusted, but then the filter's transfer characteristic was altered significantly from that predicted because the gyrator's Q does not depend only on the output conductances of the gyrator transistors but also on the channel delay and the external resonator (loading capacitance). Lower-order filters might work because the external terminations would provide sufficient loading of the gyrators to make them stable, but higher-order filters would have internal nodes that did not get sufficient loading to make the filters stable, unless the resonance frequencies were very low compared to the transit frequencies f
T
of the devices in the gyrator cores or every gyrator were given a separate Q-tuning circuit, which would be impractical.
German Patent Application No. 199 580 96-0 filed on Dec. 2, 1999, by S. Mattisson for “High-Q Gyrator Structures”, describes a method for enhancing the quality factor of gyrator resonators by balancing phase-lag and phase-lead terms in the gyrator structure. This works very well for parallel resonators having comparatively narrow bandwidths that emulate parallel LC resonators. Nevertheless, it is desirable for general-purpose impedance invertors to have wide bandwidths, and the job of balancing terms is complicated by the capacitances present at the gyrator terminals. In general, it is difficult completely to “absorb” those capacitances in the impedances connected to the gyrator without violating other design constraints, ultimately resulting in the gyrator's having excess phase lag. Such phase lag in a wide bandwidth gyrator manifests itself as a signal loss, with the result that the total insertion loss of the filter an become excessive. This patent application is expressly incorporated in this application by reference.
Patent Application No. 9916808.0 filed in the United Kingdom on Jul. 16, 1999, by S. Mattisson for “Integrated Circuit” describes a method of widening the us

Affiliated with

Also associated with

Rating

Phase-compensated impedance converter does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Phase-compensated impedance converter, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Phase-compensated impedance converter will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2929972