The present invention relates to solid state radio frequency switches. More particularly, the present invention relates to a convenient approach to obtain bias and other auxiliary voltages from switch control signals or voltages.
Radio frequency (RF) switches are important building blocks in many wired and wireless communication systems. RF switches are found in many different communication devices such as cellular telephones, wireless pagers, wireless infrastructure equipment, satellite communications equipment, and cable television equipment. As is well known, the performance of RF switches may be characterized by one of any number operating performance parameters including insertion loss and switch isolation. Performance parameters are often tightly coupled, and any one parameter can be emphasized in the design of RF switch components at the expense of others. Other characteristics that are important in RF switch design include ease and degree (or level) of integration of the RF switch, complexity, yield, return loss and, of course, cost of manufacture.
As further shown, several transistors M51, M52, M53, and M54 are arranged to effect RF communication between RFC 501 and RF1502, or between RFC 501 and RFC2503. Specifically, M51 is arranged between RFC 501 and RF1502, M52 is arranged between RF1502 and ground, M53 is arranged between RFC 501 and RF2503, and M54 is arranged between RF2503 and ground. Each of the transistors M51-M54 includes by-pass resistors (which are not labeled with reference numerals) connected between respective drain and source terminals.
Two control voltages VC1 and VC2 applied, respectively, to the gates of M51 and M53 control which path (RFC to RF1 or RFC to RF2) will be taken by an RF AC signal input at RFC 501. In the configuration shown, VC1 is 3.3V, which turns M51 on. VC2 is 0V, which turns M53 off. In this configuration the RF path is configured to be RFC to RF1. M52 and M54 operate to enable either an isolation branch or a shunt branch depending on which path (RFC to RF1 or RFC to RF2) is selected. That is, when VC1 is high (3.3V), VC1B (the control voltage applied to the gate of M52) is controlled to be low, e.g., 0V. With VC1B low, M52 is off thereby isolating the path between RFC 501 and RF1502. Meanwhile, VC2B is set high or equal to VC1 thus turning M54 on and creating an AC signal shunt path between output node RF2503 and ground to ensure that no signal (or very little) is present at RF2503, when RF2503 is not selected to output the AC signal received at RFC 501. The several control voltages VC1, VC1B, VC2, and VC2B are applied via respective resistors (which are also not labeled with reference numerals).
For high power operation for the switch shown in
For a depletion pHEMT device, threshold voltage (Vth) is about −1V. Due to the relatively large leakage current in a pHEMT device, two back-to-back diodes 520, 521 form a Kirchoff Voltage Law (KVL) node. Specifically, with VC1=3.3 voltage, the drop across diode 520 is approximately 0.7V, which causes a voltage of 2.6V to be present at node 525. With a voltage of 2.6V at node 525, Vgs for M52 (which is reversed biased) is 0V−2.6V, or −2.6V. The point here is that as a result of the leakage current, node 525 is set at 2.6V which is suitable for handling high power RF switch operations, and as a result, no auxiliary biasing is needed to support a high power RF switch implemented with pHEMT devices.
Unlike pHEMT device-based RF switches, silicon-based RF switches permit much less leakage current. As such, silicon-based RF switches operating in high power scenarios require special biasing circuitry and voltages to operate properly. There is accordingly a need to provide cost effective ways of providing such biasing circuitry and voltages in silicon-based RF switching devices.
Embodiments of the present invention provide an auxiliary voltage generating unit for a radio frequency switch including a first input and a second input respectively configured to receive a first control signal and a second control signal, wherein the first control signal and the second control signal are configured to control which one of a plurality of paths in the radio frequency switch is enabled, and at least one output, configured to output an auxiliary voltage, derived from at least one of the first control signal or the second control signal, that is used to operate the radio frequency switch. The auxiliary voltage may be a bias voltage and/or a voltage used to power an inverter used to enable a selected branch as an isolation branch or shunt branch.
The present invention will become more readily apparent to those ordinarily skilled in the art after reviewing the following detailed description and accompanying drawings, in which:
Reference is now made to
In switch 100, as in switch 500, several transistors are employed to control which path is enabled and thereby permit an RF signal to pass. In the case of switch 100, transistors M11, M13 and M15 are employed as the main path controllers, while transistors M12, M14 and M16 are controlled to isolate a given path or to enable an AC shunt path for non-selected pathways. Unlike the switch 500 in
The following table shows the control voltages VC1, VC2 and VC3 that are employed to control RF switch 100. Corresponding control signals VC1B, VC2B, and VC3B are arranged to be the inverse of VC1, VC2, and VC3 by passing VC1, VC2, and VC3 through appropriate inverters 150.
More specifically, to have an RF signal pass from RFC 101 to RF1102, for example, VC1 is set high and VC2 and VC3 are set low. To have an RF signal pass from RFC 101 to RF2103, VC2 is set high and VC1 and VC3 are set low. And, to have an RF signal pass from RFC 101 to RF3104, VC3 is set to high and VC2 and VC3 are set low.
In silicon based transistors, gate leakage current is typically very small. Thus, the voltage at node 125 (shown in the top path of the switch 100 between RFC 101 and RF1102) will almost always be close to 0V since, e.g., the bypass resistor disposed across M12 is connected to ground via a DC blocking capacitor. (Node 125 is used as an example, and those skilled in the art will appreciate that the discussion herein applies equally to corresponding nodes in the other branches of the switch 100.) However, if node 125 is maintained at 0V DC, the switch does not operate well in high power scenarios. Just as in the pHEMT implementation of
To obtain the desired voltage at node 125 and support high power operations of the switch 100, a separate voltage source is employed to apply a bias signal, namely “VBIAS,” to multiple nodes in each branch and path of switch 100. Taking the top branch or path as an example, VBIAS is applied between RFC 101 and M11, at node 125, and between M12 and ground. VBIAS may be on the order of 1.6V or higher. When VBIAS is applied in this manner, the entire segment of the switch, including its respective path, isolation and shunt branches are biased to the level of the applied VBIAS voltage. As a result, the silicon-based RF switch 100 can properly operate in high power scenarios.
In addition (and not shown in
In accordance with an embodiment of the present invention, and as further shown in
The diodes 280 themselves may be formed any number of ways including with stand alone p-n junctions, NMOS or PMOS transistors connected to operate as diodes, or bipolar transistors connected to operate as diodes, and transistors 281-283 can be native MOS or standard MOS to enhance sourcing current capability. Such devices can be formed in a same semiconductor substrate in which the switch 100 is fabricated, resulting in reduced manufacturing costs.
Based on the forgoing, those skilled in the art will appreciate that embodiments of the present invention enable a silicon-based RF switch that is capable of operating in high power scenarios without the need for external voltages to obtain VBIAS and/or VREG voltages for proper operation. In accordance with embodiments of the present invention, VBIAS and VREG are obtained directly from control signals (e.g., VC1, VC2, VCn) that are used to control which path of the switch an RF signal is to take from an RF common or input node to an output node. VC1, VC2, VCn, for instance, are pins that might be found on an RF switch implemented as an integrated circuit. Without the need for yet more pins to supply VBIAS and/or VREG, the packaging and overall circuit layout of such an integrated circuit can be simplified, resulting in reduced cost.
As those skilled in the art will appreciate, the auxiliary voltages may be a VBIAS voltage that is applied to maintain certain nodes along respective branches of the RF switch at a voltage level sufficient to sustain high power operation, or a VREG voltage that may be used to power an inverter which is used to invert control signals to enable isolation or shunt branches within the RF switch. In either case, by generating such auxiliary voltages using control signals, it is possible to simplify the design of, e.g., a silicon-based RF switch. That is, no separate, external-supplied VBIAS and VREG are needed.
While the invention has been described in terms of what is presently considered to be the most practical and preferred embodiments, it is to be understood that the invention needs not to be limited to the above embodiments. On the contrary, it is intended to cover various modifications and similar arrangements included within the spirit and scope of the appended claims which are to be accorded with the broadest interpretation so as to encompass all such modifications and similar structures.