How to Implement a High Sensitivity Spectrophotometric Sensing Circuit
Rising concerns over water and air quality have pushed designers of laboratory and analytical spectrophotometry instrumentation to quantitatively analyze for increasingly subtle contaminants or discoloration in gases or liquids. However, the increasingly minute levels require equally sensitive detection methods to measure the intensity of absorbed or deflected light after it passes through a sample solution.
The challenge for designers is to be able to design low noise and ultra-low current front-end electronics that minimize measurement interference with the sensing devices. Standard transimpedance amplifier (TIA) circuits with front-end photodiodes are not precise enough to satisfy the increasingly high sensitivity requirements of the analytical spectrophotometers.
For many designers, the best path is to simply tweak existing circuits. Using this design technique lowers overall cost while ensuring maximum chance of a successful design.
This article discusses the requirements of a TIA circuit for a high-precision, low-current photodiode. To accommodate extremely low photodiode currents, it introduces the critical elements of the signal chain, including an Analog Devices ADA4530-1ARZ-R7 low noise front-end amplifier and an AD7172-2BRUZ high-precision analog-to-digital converter (ADC), along with optimum layout techniques. It then describes how to get a design off the ground using a reference design that combines well-matched elements in a practical configuration.
Spectrophotometry uses quantitative analysis in various fields such as chemistry, biochemistry, physics, chemical, and material engineering. The technique measures the absorption or reflection light incident upon a substance, in this case a substance suspended in water. The measurement fixture senses the light intensity as a beam passes through a sample solution. A typical spectrophotometer comprises a light source, a collimator, a monochromator, a wavelength selector, a cuvette for the sample solution, a photoelectric detector, and a digital display or a meter (Figure 1).
Figure 1: A spectrophotometer takes advantage of the fact that every chemical compound is distinguishable by how it absorbs, transmits, or reflects a specific range of light wavelengths. (Image source: Chemistry LibreTexts)
In Figure 1, the collimator, monochromator, and wavelength selector produce the desired wavelength from a light source. The collimator directs a straight beam of light to the monochromator. The monochromator or prism creates several wavelengths or a light spectrum. The wavelength selector (slit) filters the light signal down to a narrow selected band of wavelengths. The resulting incident light signal (Io) then impinges upon a sample solution held in a cuvette, a straight-sided, optically clear container for holding liquid samples.
After the desired light wavelength passes through the cuvette’s sample solution, the transmitted light (It) is detected by a photodetector, which senses the number of emergent photons. The signal is further processed to eventually go to a digital display.
Every chemical compound absorbs, transmits, or reflects a specific range of light wavelengths. Spectrophotometry equipment measures the type and amount of a chemical substance through absorption or transmission by measuring the light intensity output of the sample solution.
There are two different types of spectrophotometers, each dependent on the monochromator’s wavelength range.
The ultraviolet (UV) visible spectrophotometer with a wavelength range split into two: 185 to 400 nanometers (nm) and the visible range of 400 to 700 nm.
The infrared (IR) spectrophotometer, with a wavelength range of 700 to 15000 nm.
Applications for spectrophotometry abound. In biochemistry, for example, spectrophotometry is used to analyze catalyzed enzyme reactions. The technique is also used to clinically examine blood or tissue. Other spectrophotometry variations include atomic emission spectrophotometry and atomic absorption spectrophotometry.
A classic photodetector stage uses a photosensor such as a silicon photodiode or photomultipler to convert light to a small current. An operational amplifier (op amp) then follows the optical sensor to convert the small sensor current to a usable voltage. Simply put, this describes a basic TIA.
The critical components in a TIA circuit are the photodiode, a low-input bias current op amp, a feedback resistor (RF), and a stabilizing feedback capacitor (CF) (Figure 2).
Figure 2: A basic TIA converts a small sensor current (IPD) from a photodiode to a usable voltage. The critical components are the photodiode (DPD), a low-input bias current op amp, a feedback resistor (RF), and a stabilizing feedback capacitor (CF). (Image source: Bonnie Baker)
In Figure 2, the photodiode is selected to sense either the UV visible or IR wavelength ranges. The op amp has high impedance inputs with minimal input bias current, with ranges of tens of picoamperes to tens of femtoamperes (fA). RF can range from hundreds of kiloohms (kΩ) to tens of gigaohms (GΩ), and is sufficiently high to convert the photodiode’s current (IPD) to the full output voltage range of the amplifier. CF, whose value depends on the relationship between the amplifier’s bandwidth, input capacitance, and the parasitic photodiode capacitance, establishes the phase margin of the TIA.
The primary challenge in TIA design is to ensure circuit stability. This analysis will evaluate the transfer function of the TIA with a Bode plot.
A typical TIA circuit is shown (Figure 3). The circuit’s stability depends on achieving a balance between the amplifier’s gain and bandwidth characteristics (AOL(jw)), the circuit’s two resistors, and six capacitors.
Figure 3: In a TIA photosensing circuit model, stability requires balancing amplifier gain and bandwidth characteristics (AOL(jw)), the circuit’s two resistors, and six capacitors. (Image source: Bonnie Baker)
In Figure 3, the photodiode model has an ideal diode with the light induced current source (IPD), parasitic junction capacitance (CPD), and parasitic junction impedance (RPD). The other parasitic capacitances in the TIA that impact the circuit stability are the amplifier’s common-mode input capacitance (CCM), the differential input capacitance (CDM), and the feedback resistor’s parasitic capacitance (CRF) (Figure 4).
Figure 4: Definition of the resistances and capacitances in the TIA circuit per the model in Figure 3. (Image source: Bonnie Baker)
The frequency domain transfer function of the TIA is given as per Equation 1:
AOL(jw) is the open loop gain of the amplifier over frequency
β is the system feedback factor, equaling 1/(1 + ZIN/ZF) where:
ZIN is the distributed input impedance and equal to RPD || jw(CPD + CCM + CDIFF)
ZF is the distributed feedback impedance and equal to RF || jw(CRF + CF)
The Bode plot helps determine the circuit’s stability. The appropriate Bode plot for this design has the amplifier’s open-loop gain and the 1/β curve. System elements determining the noise gain (1/β) frequency response are the photodiode parasitics and the op amp’s input impedance (ZIN), as well as the components in the amplifier’s feedback loop (RF, CRF, and CF) (Figure 5).
Figure 5: The closure rate between the open loop gain frequency response and feedback gain reciprocal (1/β ) is 20 decibels (dB)/decade. (Image source: Bonnie Baker)
In Figure 5, the green curve shows the closed loop gain of the TIA and the teal curve shows the open loop gain performance of the ADA4530-1. In the closed loop TIA gain curve, the gain at DC is equal to the non-inverting gain of the amplifier circuit, with gain equaling 1 + RF/RPD. The first change in frequency with this curve occurs at the first zero (fz), which depends on the feedback network. The second change in frequency of the TIA closed loop gain curve occurs at the first pole (fP), which depends on the photodiode parasitics, amplifier parasitics, and the feedback elements. This gain curve theoretically flattens at a final gain of 1 + (CPD + CCM + CDIFF)/CF. To calculate fZand fP, Equations 2 and 3 are used:
What’s interesting in this circuit is where the AOL(jw) curve intersects the 1/β curve. The closure rate between these two curves determines the system’s phase margin, and in turn, predicts the stability.
For instance, the closure rate of the two curves in Figure 5 is 20 dB/decade. The amplifier is contributing approximately a –90 degree phase shift, and the feedback factor is contributing approximately a zero degree phase shift. By adding the 1/β phase shift from the AOL(jw) phase shift, the system’s phase shift is –90 degrees and the phase margin is 90 degrees, resulting in a stable system. If the closure rate of these two curves is 40 dB/decade, indicating a phase shift of –180 degrees and a phase margin of zero degrees, the circuit will oscillate or ring with a step function input.
Two ways to correct circuit instability is to add a feedback capacitor, CF, or change the amplifier to have a different AOL frequency response, or different input capacitances.
A conservative calculation that allows variation in amplifier bandwidth and input capacitance, as well as the feedback resistor value, places the system’s pole of 1/β at half the frequency where the two curves intersect. This calculation for CF is shown in Equation 4:
Where fGBW is the amplifier’s gain-bandwidth product. Also, Equation 4 produces a system phase margin of 65 degrees.
For example, Analog Devices’ ADA4530-1ARZ-R7 fA input bias current electrometer amplifier has a maximum input bias current of ±20 fA, a 50 microvolt (µV) input offset voltage, and an fGBW of 1 megahertz (MHz), with CCM plus CDIFF equaling 8 picofarads (pF). The components outside the amplifier—RF, CRF, and CPD—are 10 GΩ, 5 pF, and 1 pF, respectively.
Proof of concept: spectrophotometer detector
As mentioned earlier, a photodiode/precision amplifier detects and converts incident photons on the photodiode to a usable voltage. A high-resolution ADC then converts the amplifier’s output voltage to a digital representation. The functional schematic for this is shown in Figure 6. The spectrophotometer detector stage must measure photodiode currents in the femtoampere range with a precision analog front-end. The TIA’s input bias current specifications must comply with this low input bias current requirement.
Figure 6: Spectrophotometer femtoampere TIA detector circuit based on the ADA4530-1ARZ-R7 femtoampere input bias current electrometer amplifier uses a low leakage mezzanine board (left) connected to a data acquisition board (right). (Image source: Bonnie Baker)
The TIA circuit shown uses two boards: a low leakage mezzanine board mated to a data acquisition board. The mezzanine board contains the photodiode (DPD), the ADA4530-1 ultra low input bias current op amp, the extraordinarily high feedback resistor (a 10 GΩ glass resistor), and a feedback capacitor (CF) to form a basic TIA circuit.
The appropriate input devices for this ultra high sensitivity analog front-end are photodiodes or photomultiplier tube sensors. The sensing diode (DPD) spans across the differential input pins of the ADA4530-1. An integrated guard buffer in the ADA4530-1 ensures its ±20 fA input bias current remains low by isolating the input pins from the pc board leakage.
For the test performed in this article, the mezzanine board (EVAL-CN0407-1-SDPZ) is a low leakage board based on a hybrid FR-4 and Rogers 4350B laminate. The outside layers are ceramic (Rogers 4350B), and the inside layer is a standard glass epoxy laminate (FR-4). Compared to glass or epoxy materials, the Rogers 4350B material is a better insulator (Figure 7).
Figure 7: The low leakage mezzanine board used in this TIA setup is a hybrid FR-4 and Rogers 4350B laminate. (Image source: Analog Devices)
In Figure 7, the Rogers 4350B material also minimizes current leakage, and as compared to glass or epoxy dielectrics, has much shorter dielectric relaxation times.
ADC and voltage reference
The data acquisition board has an Analog Devices AD7172-2 ADC, a power supply module, the ADC’s reference voltage, and an isolated digital interface. The ADC is a 24-bit Ʃ-Δ ADC that produces 24 noise-free bits at a conversion rate of 5 samples per second (SPS).
The output voltage range of the mezzanine board is ±5 volts. With Analog Devices’ ADR4525BRZ-R7 2.5 volt voltage reference, the AD7172-2 ADC’s input range is ±2.5 volts. The 10 kΩ/10 kΩ matched resistor divider attenuates the output of the mezzanine board by a factor of two. To minimize ADC offset errors, an Analog Devices ADG1419BRMZ-REEL7 analog single-pole/double-throw (SPDT) switch shorts the input of the resistor divider to ground. This configuration allows the removal of the measured the ADC and resistor divider offset error. The ADA4530-1’s own circuitry generates the remaining offset.
The power management portion of the spectrophotometer femtoamp detector stage powers all components on the mezzanine and data acquisition boards. The power management section, on the data acquisition board, derives its power from a 9 volt external DC power supply (Figure 8).
LDOs). (Image source: Analog Devices).png" alt="Figure 8: Using an external 9 volt input, the power section of the spectrophotometer femtoamp detector powers all components on the mezzanine and data acquisition boards using Analog Devices low dropout regulators (LDOs). (Image source: Analog Devices).png" style="width: 642px; height: 397px;" width="642" vspace="0" height="397" border="0"/>
Figure 8: Using an external 9 volt input, the power section of the spectrophotometer femtoamp detector powers all components on the mezzanine and data acquisition boards using Analog Devices low dropout regulators (LDOs). (Image source: Analog Devices)
The input circuitry from the 9 volt external input to the board’s power ICs includes protection against overvoltage transients and reverse voltage. Three Analog Devices ADP7118ACPZN-R7 low-noise, LDO linear regulators generate 5 volts for the ADA4530-1 amplifier, 2.5 volts for the AD7172-2 ADC analog front-end, and 3.3 volts for the digital input/output lines and the Analog Devices ADUM3151BRSZ-RL7digital isolators.
Testing the spectrophotometer detector circuit
The mezzanine board rides on top of the data acquisition board as shown in Figure 9.
Figure 9: The combination of the mezzanine and data acquisition pc boards before a shield is placed around the mezzanine board. (Image source: Analog Devices)
In Figure 9, the mezzanine board is shown with the shield removed. Once in place, the shield prevents interference at the input stage of the ADA4530-1 amplifier.
To start testing, the 9 volt supply needs to be connected and the EVAL-CN0407-SDPZ evaluation software downloaded from the Circuit Evaluation & Test section of Analog Devices’ support site.
Once the software is up and running, the board is then configured to test the ADC noise. For best noise performance, select the lowest acceptable sampling rate. For example, the system noise when sampling at 0.83 SPS for 120 minutes produces a root mean square (rms) noise of 1.4 fA with a DC value of −150 attoamperes (aA) (Figure 10).
Figure 10: For best noise performance of the femtoampere measurement system, select the lowest acceptable sampling rate. For example, shown is the system noise when sampling at 0.83 SPS for 120 minutes. This produces a root mean square (rms) noise of 1.4 fA, with a DC value of −150 aA. (Image source: Analog Devices)
The thermal noise from the 10 GΩ resistor, equaling 12.87 µV/√Hz, will dominate system noise. To counteract this, the oversampling capability of the ADC can filter the higher frequency noise from the results.
Spectrophotometry instrumentation quantitatively analyzes for subtle contaminants or discoloration in gases or liquids. The challenge for designers is to be able to design low noise and ultra-low current front-end electronics that minimize measurement interference with the sensing devices.
In pursuit of a viable spectrophotometry solution, it has been demonstrated that a TIA configuration, comprising an ADA4530-1 femtoamp amplifier and a 24-bit AD7172-2 Ʃ-Δ ADC, can be used to create a high-precision, robust solution. Innovative layout and board manufacturing techniques help realize the final solution and produce a low noise result.
1.The content, data, charts, etc. of this article come from network reference or other public materials, and the copyright belongs to the original author and the original published source. If the copyright owner has any objection to the quotation of this article, please contact ICZOOM "marketing(at)iczoom.com" and we will deal with it in a timely manner.
2.The quotes in this article are for readers' learning exchange only, and do not involve commercial purposes.
3.The content of this paper only represents the author's point of view. ICZOOM cannot gurarante and assure the accuracy, reliability or integrity of the content. The decision or behavior made by readers after reading this article is based on their own will and independent judgment. Please clarify the relevant results before reading this article.
4.Please contact ICZOOM "marketing(at)iczoom.com" with the reason of reproducing if you want to reproduce the articles that ICZOOM owns the copyright. Without permission to reproduce, ICZOOM will reserve the right to pursue the legal liability.
5. If there is any inconsistency between the English and Chinese versions, the Chinese version shall prevail.
ICZOOM has the final right to interpret this statement.