No. The number of data bits is different and it is a two wire interface. The SDI protocol has to be implemented via software.

It contains an integrating sigma-delta converter with selectable over sampling ratio. Also over sampling frequency is selectable for 2MHz or 4 MHz. To meet full accuracy as specified in data sheet, 2 MHZ gas to be used!

Averaging factors from 1 to 1024 can be set. This enables sampling rates from 8(16) kHz down to 2 Hz. If the active offset-suppression is switched off, the ASIC will measure with 32(64) kHz with the full 16-bit-resolution.

Active offset compensation of each measuring value provides an offset below 1 µV. This compensation is done with a rate of 20 kHz resulting also in a negligibly low offset temperature drift. Another positive effect of this active compensation is the complete elimination of the 1/f noise.

It is possible with a high level power supply (approx. 5 mA). In this case, the signals of the SDI-bus have to be level shifted (by opto-coupler).

Yes, it is a low side measurement system and measures small voltages referred to ground. For measurement of large voltages a voltage divider is used to scale down the 42V for example to 30 mV.

Both products are fully available.

Yes, this software can be provided in C language.

With the internal comparator a limit value can be set and observed. If it is exceeded, an interrupt is provided to the µC.

Yes, if there is no external clock the ASIC will work automatically with the internal oscillator (125 kHz). All measurement ranges and additional functions are fully operative, only the measuring frequency is lower by the factor 4 MHz/125 kHz.

The internal resistance of the battery can be determined by statistical or systematic fluctuations of the power drawn from the battery (caused by the periodic current requirement of the injection equipment, for example). The dual-mode will allow voltage and power measurement quasi simultaneously. The internal resistance can be calculated directly from current and voltage fluctuations. Due to noise in voltage and current, the calculations may cause major uncertainties of the resistance value. Due to the high sampling rate of the ASIC precise calculation of the real dynamic internal resistance is possible. As the measured data contains also frequency information the frequency dependence of the internal resistance is known and can be normalized for example to idle frequency.

Under standard conditions, the current consumption is typically 2.9 mA. Main sink is the analogue part due to the required extremely low noise (2.4 mA). The ASIC can also be operated without external clock. In this case, the internal clock generator will start automatically with 250 kHz and the current consumption will drop to approx. 2.5 mA.

The internal sampling rate is maximum 64 kHz. Four cycles are necessary for the continuous digital offset compensation. Therefore, the maximum measuring rate is 16 kHz. The offset compensation can be switched off to reach an actual sampling rate of 64 kHz. However, a low-pass filter in front of the programmable amplifier limits the bandwidth to 20 kHz. With an amplification of 1 the amplifier and filter is eliminated from the measurement path and the signal is directly connected to the AD converter. A version for higher sampling rates is not planned. Please note that a over sampling frequency of 2 MHz is to be chosen to reach the measurement accuracy as specified in the data sheet.

AC-measurements are possible as input signals above and below ground can be measured. Measurement at the inputs ETR, ETS, Vbat and RSHH is done differentially to RSHL. If you want to measure large voltages, a voltage divider is required to scale down the voltage to +/- 30 mV for example at gain 24.

The absolute precision of the measurement is better than the tolerance of the shunt as both the current as well as voltage and temperature can be calibrated separately for each amplification. The combination chip+shunt can be calibrated for absolute value as well as for minimum temperature coefficient. It is important that the thermoelectric voltage of the resistance is negligibly small versus copper, thus Manganin® is suited ideally.

The separation between analogue and digital part has been a very important aspect for the design of the AS8500 / AS8501. They are spatially separated and also have their own supply voltages and mass pins. There is no cross talk noticeable between analog and digital part of the ASIC. Nevertheless, there is no electrical isolation between analogue and digital part. Both VSS pins have to be connected low-ohmic externally on the PCB. With reference to coupling of high-frequency clock signals, both areas are decoupled so well by a suitable design that the 100 µOhm resistor will allow the resolution of 1 mA DC-currents.

Integration into the CAN world has to be realized via µcontroller that is required anyway.

The products contain three important features, making them an ideal data acquisition system and making life easier with reference to calibration: 1. almost zero offset, 2. very high linearity, 3. extremely low noise. Thus, a complete calibration will only require multiplication with a calibration factor. The IC’s have a read-only memory where for every single measuring channel an 11-bit word can be filed as calibration factor. Starting up, the content of this register will be loaded into the RAM area of the CAR register. The selected value will be loaded in CRA and CRB, multiplied with each value delivered by the converter and made available through SDI. If the stored calibration factor turns out not to be suitable or not correct (any more) or is a default setting (for AS8500), the CAR register may be overwritten after each power-up via serial port; thus, specific adjustment are always possible. Furthermore, fine calibration by the external Microcontroller is of course also possible. The source text in C for the general serial communication between ASIC and µC is available.

For a 20 channel temperature measurement for instance, only five IC’s will be needed. Four input channels per component can be used. They can easily be multiplexed via serial clock resulting in just one µ processor. Such a compact solution can easily be realized in matchbox size. Most probably, the total size will considerably be determined by bonding of the 20 temperature sensors.

In the single mode, the maximum measuring frequency is 16 kHz, in the dual mode it is maximum 8 kHz for both. Each channel will measure with 4 kHz due to the change-over during offset compensation. First, the values have to be re-stored again in the dual mode, whereas the last value is still present in the single mode.

No, the RSHH channel is mandatory, the second channel can be freely chosen

Common mode input range is between -120mV and +1V.

This is useful for high temperature operation to account for the higher settling time of the measurement path.

For fast dual channel measurement with alternating ABAB.. measurement there is no internal averaging possible. However, in dual mode also sequential measurements from 1 to 16 for each channel are possible. This is selectable by CRS register setting. In that case the averaging works up to the selected number.

This is set by DSP design and can’t be altered