MAX8698CEWO+T
Product Specifications
MAX8698CEWO+T
Analog Devices — Analog & Mixed-Signal ICs
Product Overview
The MAX8698CEWO+T from Analog Devices is a high-performance analog / mixed-signal IC providing precision signal acquisition, conversion, and conditioning. Analog Devices (ADI) is a world leader in high-performance analog, mixed-signal, and DSP ICs for instrumentation, healthcare, automotive, and communications.
Superb Automation partners with leading brands including Qorvo, TI, ADI, XILINX, NVIDIA, Intel, ST, MACOM, Renesas, and Coilcraft to provide complete BOM and supply-chain solutions.
Key Features
- High-resolution data conversion with excellent SNR and THD
- Wide dynamic range and high linearity across operating conditions
- Integrated voltage reference with low temperature drift
- Flexible digital interface — SPI, I2C, or parallel options
- Industrial/automotive temperature range with guaranteed specs
- Low power consumption with power-down modes
Key Specifications
| Resolution | bits |
| Channels | — |
| Sample Rate | kSPS |
| Interface | — |
| Supply Voltage | V |
| Operating Temp | °C |
| Package | — |
| INL | LSB |
| SNR | dB |
| THD | dB |
Applications
MAX8698CEWO+T: A Power Management IC Tied to One Processor, Suffix and All
The MAX8698CEWO+T is a power management IC from Analog Devices that the datasheet describes as optimised for boards built around the Samsung S5PC100 application processor, covering smart cellular phones, PDAs and internet appliances. It integrates twelve power supplies with the management functions around them: three step-down converters, nine linear regulators including an always-on rail for the real time clock domain and one for the always-alive domain, and a backup battery charger for the coin cell domain. The input range is 2.7 V to 5.5 V, which covers a single lithium cell, three nickel metal hydride cells or a 5 V input, the converters switch at up to 4 MHz, the reset output is 60 ms, and the host talks to the device over a two wire interface.
Why the suffix decides whether the board boots
The revision history of the datasheet is the interesting document for a buyer, because it shows the part number changing with the silicon. Revision 2 moved the default voltage of LDO5 to 2.8 V and LDO6 to 2.6 V and revised the power up sequence so that the two main step-down converters are controlled by the hardware enable pin unless their register values are cleared after power up. Revision 4 changed the orderable part number itself to MAX8698CEWO+T. A processor rail set that was written around revision 0 does not necessarily match a device built to revision 8, and the visible difference is a rail that comes up at the wrong voltage or a sequence that starts in the wrong order.
For anyone maintaining a board built around this device, that means the orderable number has to be quoted in full, including the C, and the datasheet revision has to be part of the purchase record. A generic MAX8698 quotation is not the same part.
4 MHz switching is what makes the solution small
The step-down converters run at up to 4 MHz, which is a 250 ns period, and the core converter at 1.2 V from a 3.7 V cell therefore runs at 32 percent duty with an 81 ns on time. With a 1 uH inductor that gives 203 mA of ripple current, and with 2.2 uH it gives 92 mA. Those numbers are the input to the inductor selection: a small inductor is possible, but not below the point where ripple current reaches the output capacitor's rating. The same arithmetic at a 5 V input gives 24 percent duty, so an adapter powered version of the board needs the inductor checked again rather than copied from the battery version.
The rails and the discharge resistors
The regulator map is three step-down converters, the first fixed at 1.8 V for memory, the other two serial programmed and pin controlled so that the processor core and its internal memory can be moved independently, plus the nine linear regulators and the charger. The converters start with the core and internal memory rails off, defaulting to 1.2 V when enabled, which is the behaviour an S5PC100 power sequence expects.
Every rail except the always-on ones carries an internal off discharge resistor of 1 kohm, plus or minus 30 percent, per the datasheet table. On a 1.2 V rail that is 1.2 mA of discharge current while the rail is disabled, and a 10 uF output capacitor decays with a 10 ms time constant, reaching one percent in about 46 ms. That is short enough to guarantee a clean restart, and it is also a small standing load that has to be accounted for in a low power state.
| Parameter | Datasheet value | What it decides |
|---|---|---|
| Rails | Three step-down, nine linear, coin cell charger | One device for the whole processor rail set |
| Input range | 2.7 V to 5.5 V | Single lithium cell or 5 V adapter |
| Switching frequency | Up to 4 MHz, 250 ns period | Small inductors, small output capacitors |
| Core converter at 1.2 V from 3.7 V | 32 percent duty, 81 ns on time | 203 mA ripple with 1 uH, 92 mA with 2.2 uH |
| Power up sequence | Revised at revision 2, registers must be cleared | Sequence differs from early silicon |
| Default voltages | LDO5 2.8 V, LDO6 2.6 V after revision 2 | Rail set depends on datasheet revision |
| Reset output | 60 ms | Holds the processor until all rails are up |
| Off discharge | 1 kohm plus or minus 30 percent per rail | 1.2 mA on a 1.2 V rail, 10 ms time constant |
| Package and range | 42 pin, -40 C to +85 C | Handset class thermal envelope |
Two questions before anyone quotes this part
Can it be replaced by another MAX8698 version? The A, B and C versions differ in default voltages and sequencing, and the C revision is the one tied to this processor family. A replacement has to match the processor's rail and sequence requirements, which are documented per processor rather than per regulator, so the check is against the processor data sheet.
Is an obsolete processor platform worth keeping alive? For a product still in service, yes. The device is a single point of failure for a whole board, and the alternative to finding stock is a board respin. That is the conversation this part number usually starts.
The device is listed in the analog and mixed-signal IC category. Sourcing conversations for a single source regulator map start at power management cross reference, and short lead time recovery is described on BOM shortage recovery.
MAX8698CEWO+T: at 4 MHz the output ripple belongs to the capacitor's resistance, not its value
Four megahertz is one quarter of a microsecond per cycle, and at that frequency a ten microfarad output capacitor is no longer the element that sets ripple. The inductor current swings by 200 mA in the core converter, and what that swing turns into at the output is decided by the equivalent series resistance of the capacitor and by the resistance of the trace between the capacitor and the load. This block works the ripple, the standing load of the discharge network and the heat of nine linear rails in one 42 pin body.
Two hundred milliamps of inductor ripple becomes 1.6 mV, and 60 percent of it is ESR
With 1 microhenry at 4 MHz and a 1.2 V output from 3.7 V, the duty cycle is 32.4 percent and the on time is 81 ns, which gives an inductor ripple of 203 mA. Into a 10 microfarad capacitor the capacitive term is 0.63 mV, because the capacitor charges for only part of a 250 ns period, while 5 milliohm of series resistance across 203 mA contributes 1.02 mV. The two add to about 1.65 mV, so the resistance is the larger term and a capacitor with twice the value and the same resistance changes nothing. Moving to 2.2 microhenry cuts the ripple to 92 mA and the total to about 0.75 mV, which is the cheaper route to a quieter rail than a larger capacitor.
The discharge network is a standing load in every rail, awake or asleep
Each rail carries a 1 kohm discharge resistor with 30 percent tolerance, so a 1.2 V rail draws 1.2 mA, a 1.8 V rail 1.8 mA, a 2.8 V rail 2.8 mA and a 3.3 V rail 3.3 mA. Those currents flow whenever the rail is up, not only when it is being discharged, and they are the same order as the quiescent current of the converter that feeds them. On four rails that is about 9 mA of standing load, which is a measurable fraction of a handset sleep budget and is the reason the discharge function is usually disabled in the lowest power state and enabled only for a restart.
Nine linear rails in a 42 pin body put their heat in one place
A linear regulator dropping from 3.7 V to 2.8 V at 150 mA dissipates 0.135 W, and its sibling dropping to 1.8 V at the same current dissipates 0.285 W. Add the other rails and the package is carrying close to half a watt of linear dissipation in addition to the three switchers, in a 42 pin body rated to 85 C. That is why the rail assignment matters: moving a 150 mA load from a linear rail to a switching one removes more heat than any amount of copper under the package, and why a board that runs the processor at full load in a warm enclosure has to be checked at the package rather than at the battery.
| Rail | Discharge current at 1 kohm | Linear drop from 3.7 V at 150 mA | Heat |
|---|---|---|---|
| 1.2 V, switcher | 1.2 mA | Not applicable | Switcher loss only |
| 1.8 V, linear | 1.8 mA | 1.9 V | 0.285 W |
| 2.8 V, linear | 2.8 mA | 0.9 V | 0.135 W |
| 3.3 V, linear | 3.3 mA | 0.4 V | 0.060 W |
The rail set and the sequencing differences between versions are indexed on power management cross reference, the checks before a different revision goes into a qualified board are on how to verify a drop in replacement, and the passive line around the converter is quoted through one stop BOM sourcing.
Request Quote — In Stock, 100% New & Original, Full Warranty