One of the most ubiquitous ICs in electronic systems is the analog switch found in both analog and digital systems. Its specific applications include battery-based equipment, communications systems, medical equipment, test equipment, and audio and video switching. General-purpose applications include sample and hold circuits, digital filters, op-amp gain switching, analog signal switching, multiplexing, and integrator reset circuits.
Some of the newest precision analog switches can operate from ±1.5V to ±6V supplies, such as Intersil's ISL84516 and ISL84517. They are housed in SOT-23 packages and are well suited to portable battery-powered equipment — thanks to the low operating supply voltage (±1.5V), low power consumption (350 µW), and low leakage currents (2 nA max). Both switches are single-pole single-throw (SPST) types; the ISL84516 has normally open (NO) contacts while the ISL84517 has normally closed (NC). High-frequency applications also benefit from the wide bandwidth and very high off isolation. The figure shows how to configure the device for overvoltage protection.
Break-before-make operation makes them ideal for custom multiplexer applications. Additionally, excellent RON flatness maintains signal fidelity over the whole input range, while micro-packaging alleviates board space limitations.
Low leakage currents (1 nA max) lengthen hold times in sample-and-hold circuits, and fast switching speeds (tON = 40 ns, tOFF = 30 ns) reduce propagation delay in timing-sensitive applications. These switches offer minimal charge injection (10 pC) to maximize signal integrity.
With any CMOS device, proper power supply sequencing is required to protect the device from excessive input currents that might permanently damage the IC. All I/O pins contain ESD protection diodes from the pin to V+ and to V-. To prevent forward biasing, V+ and V- must be applied before any input signals, while input signal voltages must remain between V+ and V-. If these conditions cannot be guaranteed, then one of the following two protection methods should be employed.
Adding a 1-kW resistor in series with the input can protect logic inputs. The resistor limits the input current below the threshold that produces permanent damage, and the submicroamp input current produces an insignificant voltage drop during normal operation.
Adding a series resistor to the switch input defeats the purpose of using a low RON switch, so two small signal diodes can be added in series with the supply pins to provide overvoltage protection for all pins. These additional diodes limit the analog signal from 1V below V+ to 1V above V-. The low leakage current performance is unaffected by this approach, but the switch resistance may increase — especially at low supply voltage.
The power supplies need not be symmetrical for useful operation. As long as the total supply voltage (V+ to V-, including supply tolerances, overshoot, and noise spikes) is less than the 15V maximum supply rating and the digital input switching point remains reasonable, the ISL84516/17 functions well. The 15V maximum supply rating provides 12V systems designers with greater flexibility than switches with a 13V maximum supply voltage.
Minimum recommended supply voltage is ±1.5V. The input signal range, switching times, and on-resistance degrade at lower supply voltages, and the digital input VIL becomes negative at VS = ±2V. This family of switches is not recommended for single supply applications.
Due to the lack of a GND pin, the switching point of the digital input is referenced predominantly to V+. The digital input is CMOS-compatible at ±5V supplies, and is TTL-compatible for ±3.3V supplies.
The switching point changes by only 100 mV from 85°C to -40°C, regardless of supply voltage.
The table, on page 18, compares the characteristics of these switches with other recent switch introductions.
60A VRM and VRD 9.0/9.1 DC-DC Controller
Maxim Integrated Products has introduced the MAX1937/38/39 family of dual-phase, PWM step-down, dc-dc controllers to support CPU core supplies in desktops, notebooks, and blade servers. This family of synchronous, step-down controllers delivers load currents up to 60A. They are compliant with AMD Hammer (MAX1937), Intel VRM 9.0/9.1 (MAX1938), and AMD Athlon Mobile (MAX1939) VID code specifications. An internal DAC provides ultrahigh accuracy of ±0.75% and implements a controlled VID voltage transition to minimize under and overvoltage overshoot during VID input change. The photo, on page 19, shows the 0 to 40A step-load transient response for a MAX1938.
These controllers have an 8V to 24V input operating range. Remote sensing is available for high-output voltage accuracy. External MOSFET switches are driven by a 6V gate-drive circuit to minimize switching and crossover conduction losses and to achieve efficiency as high as 90%. The family features cycle-by-cycle current limit to ensure that the current limit is not exceeded. A crowbar circuit protects against output overvoltage.
Using Quick-PWM™ control architecture along with active load-current voltage positioning, they provide instantaneous load-step response, while programmable voltage positioning enables the IC to respond to a step load transient within 300 ns. It also enables the converter to use full transient regulation limits, reducing output capacitance requirements. In addition, 0.75% output accuracy and fast active voltage positioning further reduce output capacitor values.
The heart of the Quick-PWM core is a one-shot that sets the high-side switch on-time. This fast, low-jitter, one-shot circuitry varies the on-time in response to the input and output voltages. The high-side switch on-time is inversely proportional to the voltage applied to VCC and directly proportional to the output voltage. This algorithm results in a nearly constant switching frequency, despite the lack of a fixed frequency clock generator. The frequency selected avoids noise-sensitive regions, and the inductor ripple current operating point remains relatively constant — resulting in easy design methodology and predictable output voltage ripple.
The two phases of these controllers operate 180° out-of-phase to reduce input filtering requirements and EMI. This lowers cost and saves board space, making them ideal for cost-sensitive applications. With dual synchronized out-of-phase operation, the controller's high-side MOSFETs turn on 180° out-of-phase. The instantaneous input current peaks of both regulators do not overlap, resulting in reduced input voltage ripple and RMS ripple current. This reduces the input capacitance requirement, allowing fewer or less expensive capacitors, and reduces shielding requirements for EMI. Each phase operates with a 250-kHz switching frequency. The input and output ripple have lower amplitude, with an effective switching of 500 kHz.
To minimize crosstalk noise in the two phases, the controller's maximum duty cycle is less than 50%. For a fast transient response, a phase-overlap mode allows both phases to operate in-phase when detecting a heavy-load transient. This continues until the output voltage returns to nominal output voltage regulation value. Then, the controller returns to 180° out-of-phase operation, and a minimum current-adaptive phase-selection algorithm determines which phase to use to start the first out-of-phase cycle. Once the output voltage returns to nominal regulation value, the subsequent cycle starts with the phase having the lowest inductor current. For example, if the current-sense inputs indicate that phase 2 has lower inductor current than phase 1, the controller turns on phase 2's high-side MOSFET first when returning to normal operation.
The MAX1938EEI comes in a 28-pin QSOP. Pricing starts at $2.25 (1000 up). It is specified for -40°C to +85°C. Evaluation boards are available in VRM and VRD form factors.
Intersil, Palm Bay, Fla.
CIRCLE 348 on Reader Service Card
Maxim Integrated Products, Sunnyvale, Calif.
CIRCLE 349 on Reader Service Card