CN0582

The CN0582 is a complete solution capable of measuring data from analog sensors using the provided software or API libraries.

The system offers a variable amplitude and frequency sinusoidal signal to drive an external shaker table as a test stimulus for reading back different sensor responses.

The solution comes with evaluation software that provides full control over all the board’s functions. These functions include enabling current sources and ac- or dc-coupling, input bias adjustment, data capture in both time and frequency domain, providing a single-tone or sweeping sinusoid, and data export.

The optional software API library is compatible with many different environments, and was tested using LabVIEW, C#, and Python. The API enables customers to develop their own custom functions and complete applications.

Several key features of this solution are as follows:

• 4-channel simultaneous data acquisition
• On-board IEPE excitation
• IEPE input channel offsetting
• Signal Generator:
  • Single tone or linear sweep
  • Electronic loopback testing
  • Shaker table excitation
• CN0582 Evaluation Software
  • System configuration
  • Graphical display
  • File export (data acquisition, vibration test)

 

Table 1. Hardware Specifications

Specifications		Min	Typ	Max	Unit/Remarks
Power Requirements
Input	Voltage	4.5	5	6	V
	Current	500		600	mA
Analog Input
IEPE	Full scale	0.1m*		10	V
	Bias			13	VDC
	Current source		4		mA
4 mA to 20 mA (CH3)	Resistance			249	Ω
	Power			0.25	W
THD + N			-87	-84	dB, 3.3 Vp-p at 1 kHz
SNR			99		dB, 10 Vp-p at 20 Hz to 20 kHz
Accuracy				1	% error (at greater than 100 mV)
Analog Output
Voltage		0.5		7	V peak-to-peak
Frequency		0.02		25	kHz
THD + N			-52	-55	dB, 7 Vp-p at 20 Hz to 20 kHz
Accuracy				1	% error

*Based on PGA setting of 1.

VIBRATION SENSOR TYPES

Piezo vs. MEMS Sensor

Traditionally, piezoelectric accelerometers have been the benchmark for vibration monitoring and are still the sensor of choice in the highest sensitivity applications, or where very high accelerations are involved. Microelectromechanical systems (MEMS) sensors were previously deemed unusable due to their limited bandwidth, g-range, and noise performance across frequency.

Recent advancements in MEMS sensor technology (especially with frequency response and noise performance) make MEMS sensors a viable alternative in many CbM applications. MEMS sensors are typically smaller and lower cost, and have a frequency response that extends to dc (0 Hz). One example is the CN0532, a typical MEMS based sensor with circuitry implementing an IEPE interface.

The CN0582 is compatible with piezoelectric IEPE sensors as well as IEPE-compatible sensors based on other technology. The excitation current sources can be disabled, allowing connection to voltage output sensors.

IEPE INTERFACE

Unconditioned piezoelectric sensors have a high output impedance, making it difficult or impossible to drive any length of cable. These sensors are also sensitive to cable movements and electromagnetic interference, and require precision, high impedance signal conditioning circuitry.

IEPE refers to a standard in which the signal conditioning electronics are integrated into the sensor. The internal electronics convert the high impedance charge output into a low impedance voltage signal, making it easier to transmit. The standard also uses a single conductor to provide both power and signal, which minimizes cabling complexity.

Figure 2 shows a high level block diagram of the IEPE interface. Power is supplied via a constant current source to the sensor. The sensor then modulates the output voltage depending on the measured vibration, which allows the signal to be measured across the same line as the power source.

 

 

IEPE Excitation and Control

IEPE sensors require power in the form of a constant current source to operate properly. Aside from the constant current ranging from 2 mA to 20 mA, the source must have a compliance of 24 V to 30 V, which is at least twice the typical IEPE bias voltage of 8 V to 12 V. Figure 3 shows the IEPE high voltage, constant current source topology based on the LT3092.

 

Using the formula shown in Equation 1, the constant current source for the CN0582 is set to 4 mA.

 

where:
I_OUT is the constant current output.
R_OUT is the current output set resistance.
R_SET is the current input set resistance.

 

The current source can be enabled or disabled by an ADG5401, a high voltage, low resistance switch.

Commands sent from the software to the board toggle the corresponding channel to the desired state.

Powering the ADG5401 from the same 26 V supply as the current sources ensures that the input voltage remains within limits, while the 6.5 Ω on-resistance is small enough to have an effect on the signal level.

When the current source is disabled, the input for each channel acts as a typical high impedance voltage input analog-to-digital converter (ADC) measurement. Channel 3 features an on-board load resistor that allows the user to connect and measure the data from a 4 mA to 20 mA sensor, whose value can be computed using Equation 2:

 

where:
V_SENSOR is the converted sensor voltage.
R_LOAD is the load resistor (249 Ω).
I_SENSOR is the sensor output current.

 

Input Protection

To avoid overdriving any part of the input circuitry, transient voltage suppressor (TVS) diodes are added at the input connector. These diodes ensure that the input can withstand voltages up to 40 V, protecting the sensitive analog input circuitry from dangerous voltage levels.

AC/DC Coupling

Depending on the sensor, either ac-coupling or dc-coupling may be more appropriate. DC-coupling is applicable for sensors with a response that extends to zero Hz, and that have a small, or well-specified offset voltage. AC-coupling is more appropriate for sensors that inherently have a highpass response, and may have a large dc offset that varies over operating conditions.

Figure 4 shows the ac- and dc-coupled paths, controlled by an ADG5436 high voltage, latch-up proof, dual SPDT switch.

 

 

The switch is powered from a bipolar -5 V or 26 V supply, which accommodates typical IEPE sensor outputs, as well as voltage output sensors with 0 V to 5 V, ±5 V, and 0 V to 10 V ranges. The ac-coupling path is composed of a high pass filter, whose cutoff frequency can be calculated using Equation 3:

 

where:
fc is the high pass cutoff frequency.

IEPE Input Offsetting

IEPE signals naturally have a high dc bias and peak-to-peak values, making it difficult to interface with common smaller range converters. To overcome this challenge, a level shifting circuit is used.

Figure 5 shows the level shifter implementation based on the ADA4522. The attenuation and voltage shift are designed to utilize the full ±4.096 V ADC input range by centering the input as close as possible to zero.

 

 

The required voltage shift is calculated using Equation 4:

 

where:
V_SHIFT is the level shift voltage.
V_{IEPE_INPUT} is the input bias voltage.
V_{IEPE_SHIFTED} is 2.52 (must be equal to VOCM).
G is 0.3 (computed as RFB/RIN).

 

The level shift voltage (VSHIFT) is provided by the AD5686R digital-to-analog converter (DAC). The value of VSHIFT can be programmed using Equation 5:

 

where: V_SHIFT is the level shift voltage.
V_REF is the voltage reference of AD5686R.
Gain is the gain setting of AD5686R.
D is the digital code.
N is the number of bits of AD5686R (16-bit).

 

PROGRAMMABLE GAIN AMPLIFIER (PGA)

Aside from the input signal coupling and level shifting options, the CN0582 also provides additional gain for smaller input signals. The gain can be set for each individual analog input channel.

Figure 6 shows the PGA implementation using the LTC6910. The common-mode voltage from the other signal conditioning stages is connected to the AGND pin, which serves as the reference for amplification.

 

Depending on the gain option selected, the valid input range for the CN0582 also changes. Table 2 describes the acceptable input voltage range for each gain setting; voltages outside the valid input range result in signal distortion and clipping.

Table 2. EVAL-CN0582-USBZ PGA Settings vs. Input Voltage

PGA Settings	Maximum AC Input
1	10 Vp-p
2	5 Vp-p
5	2 Vp-p
10	1 Vp-p
20	500 mVp-p
50	200 mVp-p
100	100 mVp-p

Note: The system can measure signal amplitudes of 100 μVp-p, but the signal-to-noise ratio (SNR) is compromised. Thus, the recommended minimum input amplitude is 500 μVp-p.

FULLY DIFFERENTIAL AMPLIFIER (FDA)

Since the output signal of the PGA is a pseudo-differential signal with a common-mode of 2.5 V, it needs to be amplified and converted to a fully differential signal to maximize the use of the ADC’s full scale input range of 8.192 Vp-p.

Figure 7 shows the ADA4945 fully differential ADC driver used to convert the pseudo differential signal from the PGA to a fully-differential signal for the ADC.

 

 

The PGA output with respect to VOCM is amplified and converted to a fully differential signal by the ADA4945 with a gain shown in Equation 6:

 

In case of over or undervoltage due to the input and gain combinations, the ADA4945 clamps the output to +VS or -VS to protect the succeeding sensitive ADC input stage. 

OVERALL SIGNAL CHAIN GAIN

The total gain on the analog input path comes from the level shifter, FDA, and PGA, which can be computed using Equation 7:

 

where:
G is the overall system gain of the CN0582.
G1 is the gain of the level shift attenuation (0.3).
G2 is the gain of the FDA (2.667).
G_PGA is the programmable gain of the PGA (1, 2, 5, 10, 20, 50, 100).

 

The maximum allowable ac or dc voltage on the analog input path is directly related to the gain settings of each channel. Equation 8 shows how to calculate the dc input voltage, assuming a PGA gain setting of 1.

 

The maximum dc bias input voltage is: 

 

 

where:
VIN_{DC_MAX} is the maximum dc voltage into the CN0582.
V_{SHIFT_MAX} is the AD5686R maximum output voltage (5 V).
G1 is the gain of the level shift attenuation (0.3).
VOCM is the AD7768-4 common-mode voltage (2.52 V).

The ac input voltage can be computed using Equation 9, assuming that the PGA gain setting is 1.

 

The maximum peak-to-peak input voltage swing is: 

 

where:
VIN_{AC_MAX} is the maximum ac input voltage into the CN0582.
V_{REF_ADC} is the AD7768-4 internal voltage reference (4.096 V).
G2 is the gain of FDA (2.667)
G1 is the gain of the level shift attenuation (0.3)

 

The theoretical maximum ac input swing is 10.23 Vp-p, but the CN0582 is only tested up to 10 Vp-p to meet the headroom requirements of the amplifiers. The same logic applies to maximum dc bias, which was only tested to 13 VDC. Centering the ADC output around zero by adjusting VSHIFT implies the IEPE_SHIFTED signal is equal to VOCM.

DATA CAPTURE AND DIGITAL CONTROLLER

Analog-to-digital conversion is performed by an AD7768-4, a 4-channel 24-bit simultaneous sampling Σ-Δ ADC. It contains several power scaling, filter, and sample rate options to optimize for different constraints like noise, bandwidth, and power consumption.

The AD7768-4 acts as a source-synchronous transmitter, with each channel’s data serialized on one of four data lines. The data is received by the digital controller and streamed to the host computer over USB.

The digital controller is also responsible for configuring the analog input channels, enabling the current sources, setting the ADC modes, and setting up the waveform generator.

POWER SUPPLY ARCHITECTURE

The CN0582 is powered from a single USB 3.0 port providing 5 V at 900 mA, and several dc-to-dc converters produce the required voltage rails. The system power tree consists of boost converters followed by linear dropout regulators for the analog circuits to obtain the lowest possible noise, and buck converters for digital circuits for maximum efficiency. Figure 8 presents the complete power architecture of CN0582.

 

 

WAVEFORM GENERATION AND CONTROL

The CN0582 has an on-board waveform generator used to excite and control a shaker table, which helps in testing and calibrating IEPE sensors.

Figure 9 shows the block diagram utilizing the AD9833, a programmable waveform generator with a constant voltage output of 600 mVp-p and configurable frequency output from 0 Hz to 12.5 MHz with 28-bit resolution. The output frequency can be computed using Equation 10:

 

where:
F_OUT is the output frequency.
F_REF is the reference clock frequency.
F_REQ is the frequency register value.

 

 

 

The CN0582 frequency output is band limited to 25 kHz, which is adequate for most vibration and CbM applications. The output signal is then amplified and conditioned using the ADA4522 in a high pass filter configuration, which removes the dc offset, a requirement for many shaker table inputs.

The maximum set gain of ADA4522 can be computed using Equation 11:

 

where:
Output_Amplitude_MAX is the maximum amplitude set by the gain of the ADA4522.
Rf is 102 kΩ.
Rdiff is 7.5 kΩ.

 

The signal generator can produce 8.16 Vp-p maximum, but the graphical user interface (GUI) limits the output voltage to 7 Vp-p.

The CN0582 signal generator output amplitude is controlled by an AD5543 current output DAC, configured in a current to voltage topology as shown in Figure 10.

 

 

The amplitude can be controlled following Equation 12:

 

where:
Output_Amplitude is the modulated output amplitude.
D is the digital code value.

 

After configuring the desired output amplitude and frequency, an optional post-filtering circuit is available to reduce the broadband noise. The circuit uses the ADA4522 in a Sallen-key low pass filter configuration at the DAC's output, because of the stability and high input impedance characteristics. The cutoff frequency is set to 26 kHz, and can be computed using Equation 13:

 

The cutoff frequency is adjusted to at least twice the intended IEPE bandwidth of 10 kHz, so that negligible attenuation occurs. 

The output selector stage shown in Figure 11 uses an ADG1219 switch between filtered and unfiltered output from the amplitude controller. The selected output is then buffered to isolate the circuit from the output connector.