Amplifiers – Combined with automatic amplifier disabling switch means
Reexamination Certificate
2000-02-04
2001-10-23
Pascal, Robert (Department: 2817)
Amplifiers
Combined with automatic amplifier disabling switch means
C330S069000, C330S086000
Reexamination Certificate
active
06307432
ABSTRACT:
FIELD OF INVENTION
This invention relates to a non-inverting feedback amplifier with high disabled impedance.
BACKGROUND OF INVENTION
Many system architectures benefit from the ability to disable an amplifier when not in use. In addition to the savings in power, if the disabled amplifier can display a high output impedance, several amplifier outputs can be bussed together to provide more complex functionality. In one particular application, a video cross-point switch, arrays are built from smaller building blocks. In order to implement this array, the crosspoint building block outputs must be able to disable to a high impedance. Without this ability, impedance matching would not be possible, and power dissipation would be greatly increased. These amplifiers are often used to provide gain. A non-inverting gain of two is common, to compensate for the loss of one half the signal amplitude when driving a back-terminated load. For gains that are non-inverting, and not unity, a resistor divider is typically used to provide scaled feedback around the amplifier. This resistor divider must be connected between the output and a voltage reference, often ground. When the output amplifier is disabled (its output is made high-impedance), the feedback network still loads the output of the amplifier. Typically, the impedance of the feedback network is much less than that of the disabled amplifier's output. This is an unacceptable load on the output bus. In a non-inverting, non-unity gain configuration, the feedback network must be isolated from the output node in order to provide low disabled output impedance.
Several methods of isolating the feedback network have been investigated. In one approach an extra unity gain buffer stage is used to buffer the feedback network from the output. The buffer stage may be an operational amplifier connected as a voltage follower, but any gain of +1 stage is acceptable. The buffer stage is cascaded with the output of the main amplifier. This buffer stage must be capable of being disabled and it must be a very high quality buffer because it is directly in the signal path. This method allows a high-performance solution but at the expense of high area and power. In another application an additional disableable buffer is used to bootstrap the feedback network only when the channel is disabled. The buffer is often an operational amplifier configured as a voltage follower, but any unity gain stage is acceptable. The buffer mirrors the output voltage back to the feedback node (i.e. inverting input of the amplifier). In normal operation, this buffer is disabled. When the amplifier is disabled, this buffer is enabled. By maintaining zero volts across the feedback resistor connected to the output, the current into the output is zero for all externally applied output voltages. This results in a very large output impedance limited by the accuracy of the buffer so this buffer should be as accurate as possible.
In yet another approach a buffer is used in series with the feedback network near the output node. When enabled the buffer drives the feedback network with a replica of the output voltage, maintaining the closed-loop gain control. This buffer may be disabled when the main amplifier is disabled. It always isolates the feedback network from the output bus whether or not the channel is enabled. The buffer is often an operational amplifier configured as a unity gain amplifier, but can be any voltage follower stage. Again this buffer amplifier must be very accurate because any errors it introduces to the feedback loop will be reflected back to the main amplifier. In addition its presence in the feedback path makes the feedback loop more difficult to stabilize. Another solution is to isolate the feedback network from ground or another reference using a series switching element. In practice, a saturated bipolar transistor is often used for this case. But these transistors have large non-linear offset voltages and large switching times. The use of an FET gives somewhat better results but it has a non-linear on-resistance and is not available in bipolar fabrication processes.
SUMMARY OF THE INVENTION
It is therefore an object of this invention to provide an improved non-inverting feedback amplifier with high disabled impedance.
It is a further object of this invention to provide such an improved non-inverting feedback amplifier with high disabled impedance which uses a reference buffer amplifier which need not be extremely accurate, has less offset voltage and noise errors, requires less power because its output is virtually fixed, is disabled synchronously with the main amplifiers presenting a high-output impedance without use of a series switching element, has its parasitic capacitance isolated from the output by the feedback network, and is architecturally similar to the signal amplifier such that the two may share bias circuitry.
It is a further object of this invention to such an improved non-inverting feedback amplifier which reduces crosstalk when used in multiple-output integrated circuits.
It is a further object of this invention to provide such an improved non-inverting feedback amplifier which substantially increases disabled output impedance by a number of orders of magnitude approaching that of disabled unity gain amplifiers.
The invention results from the realization that a truly effective non-inverting feedback amplifier with high disabled impedance can be achieved using in addition to the main signal amplifier a second reference buffer which mirrors a reference voltage to the reference node of the feedback network when the reference buffer and signal amplifier are enabled and presents a high impedance between the feedback network and the reference voltage when the reference buffer and signal amplifier are disabled.
This invention features a non-inverting feedback amplifier with high disabled impedance including a disableable signal amplifier and a feedback network having an input node, a feedback node and a reference node. The signal amplifier has a first input for receiving an input signal and a second input for receiving a feedback signal from the feedback node. A reference buffer includes a reference amplifier with its output directly connected to the reference node without additional series switching elements, a first input connected to a reference voltage, and a second input connected to its own output for mirroring at its output the reference voltage at its first input when the reference buffer and signal amplifier are enabled and presenting a high impedance between the feedback network and the reference voltage when the signal amplifier and reference buffer are disabled.
In a preferred embodiment the signal amplifier may include a single amplifier circuit, it may include an operational amplifier and it may include a differential input. The feedback network may include resistor elements. The reference voltage may be ground, the reference amplifier may include a unity gain amplifier.
REFERENCES:
patent: 3628129 (1971-12-01), Riley
patent: 3629719 (1971-12-01), Heller
patent: 3885220 (1975-05-01), Fluegal
Lawas Kenneth A.
Tam Kimo Y. F.
Analog Devices Inc.
Choe Henry
Iandiorio & Teska
Pascal Robert
LandOfFree
Non-inverting feedback amplifier with high disabled impedance does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Non-inverting feedback amplifier with high disabled impedance, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Non-inverting feedback amplifier with high disabled impedance will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2603332