“From milliampere to microampere to picoampere, with the development of Electronic technology and the market’s demand for low power consumption, the current level of electronic equipment has a trend of smaller development, such as mobile phone battery standby current (10−3 A) , Photodiode dark current (10-12 A), OLED pixel current (10-12 A), etc. How to accurately measure the weak current has become an unavoidable problem. In order to pass through this “no man’s land” and measure the current at the pA level, we must step into the world with 15 decimal places (fA level).However, this uninhabited area is not so easy to set foot on, the thorny chains on the road
From milliamps to microamps to picoamps, with the development of electronic technology and the market’s demand for low power consumption, the current level of electronic devices has a trend of smaller development, such as the standby current of mobile phone batteries (10−3 A), photodiode dark current (10−12 A), OLED pixel current (10−12 A) Wait. How to accurately measure the weak current has become an unavoidable problem. In order to pass through this “no man’s land” and measure the current at the pA level, we must step into the world with 15 decimal places (fA level). However, this uninhabited area is not so easy to set foot on. The thorny fetters on the road are inevitable, and many challenges need to be overcome to successfully reach the end.
The first thing to consider is the bias current problem in the measurement. When the input is open, the reading of the ideal ammeter should be zero. However, actual ammeters have some small currents when the input is open. These currents are caused by the bias current of the active device and the leakage current flowing through the insulating material inside the instrument.
Imagine that the current to be measured is as small as 1pA, while the bias current of the operational amplifier, one of the components of the ammeter, reaches 100pA. Can we still happily measure the current to be measured? Of course not, because the signal (current to be measured) has been masked by the error current (bias current), unless we can control the bias current at the fA level.
Is it enough if the bias current is satisfied? No, this is just the first step in the Long March. There are still many issues to consider next, including many design and process details that will greatly affect the measurement results. Such as noise, dielectric absorption, leakage current, insulation, shielding, PCB materials, and even cables. How do we avoid these problems in the design?
Figure 1: Challenges faced by pA current measurement
High-sensitivity detectors require precision signal chains to support extremely low detection ranges. Although there are many challenges, we have solutions. Kayden Wang, an engineer of the technology distributor Excelpoint Shijian, introduced the ADI solution-ultra-high sensitivity Fei’an measurement platform, and answered the challenges faced in the design process.
Ultra-high sensitivity Fei’an measurement platform solution
This solution is very suitable for the use of chemical analyzers and laboratory-level instruments, which require ultra-high sensitivity analog front ends to perform signal conditioning on current output sensors such as photodiodes, photomultipliers, and Faraday tubes. Applications that can use this solution include mass spectrometry, chromatography, and coulometric analysis.
► Program features
• 500pA measuring range
• Use ADuM3151 to isolate
• Femtoamp input bias current operational amplifier ADA4530-1
• 24-bit resolution ADC AD7172-2
• Use the USB interface to connect to the PC via SDP
• Simple power supply: 9VDC input, ADP7118, ADP2442, ADP7182
• Measurement synchronization
• Trigger input/output signal
Figure 2: Block diagram of ultra-high sensitivity Fei’an measurement platform
Figure 3: Functional block diagram of Fei’an measurement system
In this scheme, one of the important considerations is the choice of low-bias current operational amplifiers. Compared with conventional op amps, the amplifier ADA4530-1 used in the scheme is a fA (10−15 Class A input bias current operational amplifier, integrated guard ring buffer is used to isolate the input pins to prevent them from being affected by the leakage current of the printed circuit board (PCB), and can reduce the number of circuit board components. Compared with similar competitive devices, the input bias current is 45 times lower, and the DC accuracy is increased by 10 times. In addition to transimpedance amplifiers suitable for this solution, it can also be used for high-impedance buffers for chemical sensors and capacitive sensors.
Figure 4: Main advantages of ADA4530-1
Challenges in design
We need to consider multiple factors that may affect the measurement results in the design, in order to maximize the superior performance of the device and meet the design index requirements.
1) Design of protection
The general practice is to surround the high-impedance node with another conductor and drive the conductor to the protection voltage (equal to or close to the potential of the high-impedance node), so that no current will flow through the insulation resistance, and a better layout produces better The performance, the smaller the change in performance over time and environmental conditions.
► Protective ring
Protect surface leakage
Remove the protective ring/wire shield/wire screen
Avoid moisture absorption
Need to be driven by an amplifier (such as a buffer) that has the same potential as the input
► Protective layer
Protect the PCB body
► Via protection
Protect the side leakage current path
High source impedance and low error requirements place unrealistically high requirements on insulation resistance. The protection technology of ADA4530-1 can reduce such requirements to a reasonable level. The principle is to surround the high-impedance conductor with another conductor (guard ring) that is driven to the same potential. If there is no voltage on the insulation resistance (between the high-impedance conductor and the guard ring), then no current will flow through it. ADA4530-1 uses protection technology internally and integrates an ultra-high-performance protection ring buffer. The output of this buffer can be used externally to simplify the realization of circuit-level protection. In order to show the implementation of the protection, the voltage buffer circuit (see Figure 6) has been modified. A conductor (VGRD) is added to the model, which completely separates the high impedance (A) node from the low impedance (B) node of different voltages. The insulation resistance is simulated with two resistances: all insulation resistance between the A conductor and the protective conductor (RSHUNT1), and all insulation resistance between the protective conductor and the B conductor (RSHUNT2). Then, the ADA4530-1 guard ring buffer drives this guard conductor (via pin 2 and pin 7) to the A terminal voltage. If the voltages of the A node and the VGRD node are exactly the same, no current will flow through the insulation resistance RSHUNT1.
Figure 5: Implementation reference of protection ring and protection layer
Figure 6: Example of protection realization principle
2) Shielding design
• Shielding helps keep the stray field away from sensitive nodes, and a shielded box can be used
• The shield accessible to the operator should be grounded to ensure safety
Figure 7: Common shielding methods
3) Treatment of pollution sources
Pollution sources are prone to form weak batteries. Connect the polluted batteries to the A and B terminals of the TIA circuit to get a simplified model (see Figure 8). Both the A terminal and the B terminal are driven to the same voltage, and an error current (IBAT T) is generated, because there is a voltage drop on the output resistance equal to the open circuit voltage of the battery, as shown in the following formula: IBATT = VBATT ÷ RBATT. This battery current flows through the feedback resistor, and is combined with the signal current and other error currents in the circuit on the feedback resistor, which affects the measurement accuracy. Generally pay attention to the influence of the following pollution sources:
• Flux residue
• Dust and other particulate accumulations
• Body oil
• Salt moisture
Figure 8: The influence of pollution sources on circuit accuracy
In this regard, Shijian engineers gave the following suggestions:
• Cleaning/cleaning PCB after assembly
• Moisture will reduce the insulation of PCB and cables
Choose the right material and measure in a controlled environment
Bake after washing to eliminate absorbed moisture
• Do not use no-clean solder paste
4) Dielectric absorption
Dielectric relaxation (also known as dielectric absorption or wetting) is a characteristic of all insulating materials, which limits the performance of electrometer circuits that need to be built to a level of several fA. Commonly used PCB sheets are industry standard FR-4 glass epoxy resin. The measurement result is shown in Figure 9. It takes 1 hour for the glass epoxy sheet to dissipate the dielectric relaxation current below 10 fA. This shows that glass epoxy sheets are not suitable for high-performance electrometer circuits.
Figure 9: Dielectric relaxation properties of glass epoxy resin
Another PCB sheet considered is Rogers 4350B. Rogers 4350B is a ceramic sheet designed for use in RF/microwave circuits. Rogers 4350B is compatible with standard PCB production technology and is widely used. The measurement result of Rogers 4350B material is shown in Figure 10. The material can dissipate the dielectric relaxation current below 1fA in less than 20 seconds.
Figure 10: Rogers 4350B dielectric relaxation performance
In view of its excellent performance, Kayden recommends using Rogers 4350B sheet with ADA4530-1 in the highest performance applications. All key characteristic measurements of the ADA4530-1 are carried out using Rogers 4350B.
At the same time, based on Shijian’s support experience, the PCB processing technology is recommended:
• Four-layer board, the thickness of the finished board is about 1.5mm
• Sheet: 1/2 layer, 3/4 layer is rogers 4350B, thickness 10mil; 2/3 layer is FR408 or Roger 4450 (note that it is not ordinary FR-4 sheet).
• Copper foil thickness: 1oz inner layer, 1.5oz outer layer
• Solder mask: double-sided, green
• Characters: double-sided, white
• Via: double-sided exposed ring
• Surface immersion gold treatment
5) Cables and connectors
• Best: use triaxial cable
There is an extra internal wire to protect the signal
• As long as the potential difference between the center conductor and the shield is very small, BNC, SMA and coaxial cables are feasible
Some RF materials (PTFE) have good low leakage and low DA properties
• Bundling the cables to reduce the effect of frictional electrification
In general, the measurement of pA-level current in this type of application is compared with other range current measurement. There are many considerations in design or process. Refer to this recommended solution and design points to greatly shorten the development cycle and help us to pass through smoothly. No-man’s land for pA current measurement.