Programmable simulation IC [zz]

How to use registers for Analog Circuits and programmable parameters.

 

Address: http://www.hqew.com/tech/sheji/410361.html

 

 

Programmable SimulationICConsiderations

Analog Circuit Design and processor, FPGA andPLDIn terms of architecture, there is a fundamental difference. The signal processing function of the analog circuit remains unchanged as long as you have set the topological layout and parameters of the circuit. Of course, some existing functional circuit blocks, such as variousFilter, You can change the clock frequency (SwitchCapacitorDesign) or adjust throughDACThe various circuit factors, such as the cutoff frequency, rolling down and other functions, are adjusted in digital mode, but these adjustments can only be counted as exceptions. Most of the designs are fixed in terms of functions and almost unchangeable in terms of their running performance.

This fixed feature is not a bad thing, and in fact it is often the same. With this fixed performance, the performance of the analog signal processing channel can also be stabilized, becoming a channel with almost no changes in the system. However, in the opinion of a simulation designer, As long as FPGA or PLD is re-programmed, or it often does not need to change the Printed Circuit Board or any device on the board, the software can be easily modified through the change, when the upper level or the circuit function is adjusted, it will naturally not be able to suppress the preference on the programmable device.

In fact, simulation designers have a choice in programming! Several manufacturers now provide analog ICs that allow you to set or modify the performance of these devices as one-time production steps (see figure 5 ), you can also create or modify the layout or performance parameters of the device after your system enters the field application. This programmability provides configurability, or even reconfiguration, or provides the ability to adapt your circuit performance to the dynamic performance of your application system. Simulation IC enables you to use the same parts in different designs and provide some protection for the security of simulation circuit design, to prevent others from exploring or estimating the security and confidentiality of your designed electrical circuits by observing active and passive devices and their interconnection methods.

You should carefully consider the causes and limitations of these programmable simulator components, as well as the horizontal effects of using some unique simulator components in the system.

AboutProgram
Like many technical terms, the word "programmable" and "programmable" have different interpretations, depending on the user and the device to which the discussion object is. Programmable Digital devices can be fixedCodeOr change the processor of the Code (Program. It can be a programmable device, and Its Circuit Interconnect is programmed by the mask before leaving the factory, orOEMProgramming in the field, such as a gate array or a logical device. It may also mean that you define some function parameters, such as bit rates, and store them in internal registers.

Some of these definitions are not applicable to analog programmability; you do not execute sequential code commands. When a programmable analog circuit is used, the simulation units can be connected by the factory or by the OEM in the expected mode, or even re-programmed on the field. Another way is to set the main parameters of the circuit to a fixed topology layout, but some of them can be set and changed according to application requirements, such as filter bandwidth or rollback. However, in this case, the signal processing channel and basic functions should remain unchanged, but only parameters should be changed.

The desire and ability to program analog circuits are not uncommon. Early analog computers had plug-ins and wiring boards, enabling engineers to use basic units.ModuleConfigure and swap sub-parts of the passive device to meet various requirements. In the field of IC-based hybrid signal processing, engineers have long used the digital/mode converter (DAC) to set analog electrical circuit parameters, especially for filters, someone has already used the analog switch to reconfigure and relay the analog signal processing channel. This reconfigurable and reconfigurable routing architecture requires careful analysis of the structure Planning, printed circuit board design, and error budget.

When you need to be programmable, it seems that only digital is the only way out. However, if you are not biased towards processor and executable code practices, you also haveVariousSelect the simulation mode for the reason. For example, filters are required for almost every type of signal processing application, and even digital filters need to simulate the nyqaist frequency filter to avoid confusion during sampling. Low-level signals usually need to be amplified beforeInverterSignal compatibility and maximumSNR(Signal-to-noise ratio) and dynamic range. Even if you can use all the digital technologies to filter and shaping signals, the module method is still a solution with lower cost and lower power consumption. Taking the peak value detector as an example, if a simulation scheme is used, it is only one or two operational amplifiers and severalDiodeBut if you use a digital method, you need to include a filter, a digital/mode converter, and a processor.

Do not mix the programmable module IC with the module ASIC and the mask programming array. Some manufacturers provide various analog unit circuit libraries that they have already created models. After connecting the unit circuit, they can meet the user's application requirements. Over the years, manufacturers and users have believed that by selecting the unit circuit in the unit circuit library, it is easier to match the required analog IC. Or, more specifically, it is better to pick an Operational Amplifier unit, a single comparator, and a multiplier unit.

However, your requirements for high-performance simulation design make it more difficult for designers to configure and produce these high-performance ICS, because there are many subtle and mutually influential factors, such as hot issues, as well as various factors arising from the design layout and handling process. Many module circuit OEMs use the unit circuit Library as the starting point of the standard product design, but they also need to use their experience to make substantial adjustments to the final unit interconnection design, in order to achieve the expected purpose of the IC.

Limited varieties provided by manufacturers
When you compare the types of basic analog components from different sources, such as operational amplifiers and inverters, you will find that they differ in architecture and performance. However, these differences can be analyzed and solved. First, we need to make a meaningful comparison, starting from a relatively basic technical specification, for example, bandwidth, stabilization time, linear errors, and performance and other major measurement parameters, then look at more subtle and critical secondary measurement parameters.

For programmable simulation IC, all manufacturers provide products based on the basic principle. Some are basic analog functions, such as amplifiers, and some include more functions that can process analog signals. First, you must carefully review your application requirements, and then look at the commercially available product models that can effectively apply to your design. Existing products may be suitable for or close to your requirements, and may not be used at all. You may also need a free array that can be configured almost for any signal channel, or a filter with a certain choice of layout and parameters.

If you start by viewing basic functionsMosTransistor. Advanced linear devicesSeriesPowered programmable simulator (epad) enables you to adjust the threshold voltage and enable it before and after the last circuit is inserted (figure 1 ).Resistance. This type of epad In the transducer front amplifier andSensorIn the signal adjustment channel, electronically controlled microPotentiometer. If your application system needs protection, you can also seal the device, or place it in a place that can neither touch nor work. However, once set, internal bias voltage can be saved. Programming level of Dual-AlD 1108e and four ald110e devices,CurrentIt is 0.1 to 3mA, and the voltage range is 1 to 3 V. It can be programmed according to the 0.1v file quantity. (Note: increasing the threshold voltage is equivalent to reducing the leakage current .)

From the perspective of basic physics, the charge injected from the outside changes the epad parameter that enters the floating gate substrate. The injected charge increases the threshold level. If your design requires two-way control, you should set two back-to-back epad, because this programming is a single function (Figure 1 (B )). In addition, epad cannot be re-programmed infinitely. Standard epad circuits generally only allow 20 to 50 re-programming. Of course, this is sufficient for many application projects. For example, this number is probably used to match the characteristics of the front amplifier with the sensor or the cycle of the factory calibration.

The advantage of using epad to replace the fine-tuning potentiometer is that epad has no moving parts and is robust in the environment, and the resolution is higher than that of 10-turn mechanical devices with a standard specification of 2000 points. In addition, you can first seal the epad and then adjust it through the software. However, you must also consider some factors. For example, the retuning cycle of epad is limited, there are still many difficulties in application, and the optional values are also limited.

The epad variety is not just a basic MOS. Manufacturers also provide operational amplifiers ald1722e and microPowerAld1726e enables you to set the starting offset voltage Vos. By setting this voltage, you can reduce or eliminate the VOS error caused by the bias, andPower SupplySome consequences of rejection ratio, common mode rejection ratio, and Vos temperature coefficient error. The starting input vos of ald1726e is 50mV at factory time. You can increase the voltage610MV. The operating voltage of an operational amplifier priced at $1 can be as low as 2 V, the standard input current is less than 0.01, the bandwidth is 200 kHz, and the conversion rate is 0.17 V/MS.

When programming for epad, use E for $499100Compile the program (figure 2 ). This programmer is operated by a Windows PC and runs programming.AlgorithmThe purpose is to shorten programming time by accelerating the stability of the threshold voltage. At the same time, this programmer also reduces the overhead to exceed its target value. Programmers have a variety of epad templates. You can also combine them with a dedicated module for $199 to adapt to a variety of dedicated epad series.

Xicor is another Programmable Operational Amplifier.SupplyManufacturer, providing a series of Numerical ControlElectronicsPotentiometer (reference 1 ). The company's x9430 dual operational amplifier can set the gain, offset, and power levels to meet the needs of different applications (Figure 3 ). By adjusting the offset, you canCalibrationSignal channels are used to reduce the effect of bias drift, because temperature changes and aging often lead to drift.

A non-volatileEEPROMTo store the set parameters. In addition to the four bytes required by the operational amplifier, the eeprom also provides12Bytes can be used for data storage unrelated to operational amplifiers. ThroughSPI(X9430) the I2C serial interface of x9438 is used to program the operation of the device. You can also use the related digital potentiometer to control the offset of the Operational Amplifier in the IC separately.

For more complex biased circuits, sumnitMicroelectronicsThe company combines analog circuits and non-volatile memory into various ICs. With these circuits, even end users can fine-tune simulation functions, such as setting threshold levels for monitoring circuits. The advantage provided by these devices is that, if necessary, users can splice a part of the device to construct a circuit with infinite variation or as a part of the monitoring circuit.

For example, Summit Microelectronics has launched an sms24 programmable monitoring IC for $7.58 (1000 in batches) to set two configuration levels. You can reset the output polarity and monitor the output polarity through the dual-line serial interface or system programmer.TimerReset, write protection optimization, manual reset, and other pin and internal block functions; second, you can set the parameters used by the voltage threshold and timing cycle. The Reset Voltage threshold ranges from 2.15 to 4.75 v.

In addition, Summit also provides a heat exchange controller, model smh4042, which can be used to change the starting point of the voltage status as needed. The basic functions required by the heat exchange controller are simple in concept, that is, the system first monitors the host power supply voltage and backend voltage levels, and then connects or disconnects the power MOS that control the power supply to the card board logic circuit. However, in practice, the difference between the platform and the platform on each power supply can be as high as 50mV. On the IC that monitors host voltage and card board voltage, there is a 1 V reference output that facilitates bus pre-charging design, a programmable slope for reliable boost and buck, and compactPCISignal switching and pin detection.

Expanding the concept of this programmable parameter, but still maintaining a fixed topological layout, Summit developed the sms44 four-pair Voltage Monitor (figure 4 ). You can set the monitoring responsibilities for each monitor. The voltage range is 0.9 to 6 V, and the timer cycle of the "Watchdog" timer can be set according to the 20mV-volume file, establishes a logical combination of various monitored voltages to generate reset and interrupt outputs. You can also set the reset pulse width. Smd1108ic adds an 8-channel, 10-Bit mode/number converter, several monitoring channels, and non-volatile memory. You can also set the upper and lower limits of the converter channels on each device. If the threshold value is crossed, the IC generates an interrupt or alarm signal.

Lattice semiconduprovides a series of simulation function blocks and design tools for designs that require configurable circuits. One type isISPThe price of the pac10 IC is less than $7 (1000 in batches). It includes four filter blocks that can be connected by a simulated routing bundle (figure 5 ). Each module has two measurement amplifiers and one output amplifier. The maximum Programmable Gain of this device is620The gain range is from zero to 160000. The function of the function is accumulation and synthesis. The IC also has a 2 V v reference voltage and non-volatile memory. For ISP pac20, an eight-bit digital/mode converter and a pair of differential comparator are added.

You can use PC-based design tools and software to design your own signal processing applications. For general functions such as continuous dual-second-order and trapezoid filters, as well as various filters of different order, you can use the macro function blocks sold on the market (figure 6 ). These filters can run at a speed of 100 kHz. You can use functional blocks on the market to design the circuit, select the device and its value, define the interconnection, and run a simulation experiment to verify whether the circuit meets your design requirements.

Once you complete the simulation experiment, you can program the IC as an independent device or a device in the circuit. Internal non-volatile memory stores the set values you selected. These devices provide available bandwidth of 650 kHz and thd of 288 at 10 kHzDBThe dynamic range is 100db.

For more complex configurations, lattice recently launched the ISP PAC80 circuit, which is a 5 V programmable device used to filter a variety of topology la s, including Butterworth, cheybchev, legensid, elliptical, Gaussian, and linear phase. This continuous time fifth-order filter runs at a speed of 500 kHz, and its frequency setting accuracy is also higher than 3.5% of the target frequency, including all the requiredResistorAnd capacitors as well as a programmable gain amplifier. All internal circuits, from input to output, are differentiated. This device will300Bias voltage and precision of MVCMOSWhen the amplifier and 1 kHz AC are used together, thd is 290db performance.

You can preset two filter configurations for ISP PAC80, and then use the JTAG interface or SPI to send a command to guide the IC to implement switching between them. With this feature, your application can quickly complete the signal acquisition mode from broadband to narrow band signal extraction mode, or adapt to the signal situation. The $149 Design Software Based on Windows enables you to select from the filter library by looking at the expected performance objectives and various limitations, you can also select by identifying your favorite topology layout. You can also define your own circuit. The design software selects appropriate configurations and device values for you, provides DC and AC simulation, and forms a document for the design.

In terms of flexibility and configurability, zetex provides the reconfigurable analog universal circuit (TRAC) series (figure 7 ). This TRAC is a simulation of digital FPGA. The trac020ch device from DC to 1 MHz has 20 Analog Processing Units, which can form signal processing functions through interconnection.

Define the functions or operation objects of each unit: open circuit, short circuit, addition, inversion, logarithm, number of records, rectifying differentiation, integration, amplification and attenuation. (Middle School Mathematics: multiplication and division can be done by adding addition and subtraction to logarithm and Cosine Functions ). TRAC uses all on-chip devices to implement most of the above operations, but some users need to provide external resistors and capacitors for implementation. trac ic stores the functions you need in the internal storage device. However, you need to plot the diagram of your signal processing function, and then use TRAC units for implementation (Fig 8 (a) and (B )).

When developing a TRAC signal processing channel, you can use a PC running Windows and a development tool from the manufacturer. The $460 full suite of tools on the simulation board has four groups of 20-unit TRAC devices and EEPROM, design software and files. With this software, you can design it, select the desired unit function from the tool frame of the software, and place it on the expected unit, and add the interconnect plug-in, select various internal devices and various internal and external connections. The next step is to simulate the performance and check the circuit performance after configuration. Finally, you can use parallel cables to download the configuration data that you think is acceptable from the PC to the TRAC register. If you are satisfied with the implementation scheme of TRAC, you can use the metal mask interconnect layer that limits the performance of high-frequency parameters to make it a fixed function, low-cost devices.

Are you sure you want to accept this?
When you use the same unprogrammed component in many parts of the design or cross-cutting in many products, you can use a programmable simulation IC in terms of flexibility, customization as required, and inventory management and outsourcing, it has some obvious advantages. Of course, some key requirements need to be carefully considered before implementation.
First, consider what kind of simulation performance can be achieved. You can get thousands of simulators from analog IC manufacturers. You can find the best devices in the first and second level technical specifications, such as AC performance, DC performance, noise, power consumption, and cost, or a combination of some technical specifications to meet your requirements. Compared with conventional simulators with these fixed functions, the programmable analog IC does not burst out of a single performance parameter or combination parameter. Programmable devices can balance various performance factors, but this balance may not be suitable for your application requirements. These devices tend to target low-frequency to medium-frequency devices. They require better performance, but not necessarily outstanding performance.
The second thing to consider is that with the programmable simulator, you have changed the course and deviated from the basic reality of the simulation IC: many basic devices, for example, both an operational amplifier and a converter have multiple replaceable circuits. The manufacturer provides some patented or top-notch devices, which are the best in terms of one or more performance parameters. However, if the manufacturer has problems with delivery, you usually need to find a similar or acceptable alternative to pin compatibility, and even have to make some concessions or compromise.
Because many systems also use patented products, there are no alternatives to some part of the design, so most designers and project managers have the idea to use as few products as possible for this one. Designers use such exclusive products only when they take the lead in the competition. Under normal design conditions, it is usually within the scope of analog signal processing or power supply functions. You should take adequate preventive measures in case of problems with the manufacturer or manufacturer's products.
When you purchase a non-programmable device, you plan to take care of the device inspection, fine-tuning, and encapsulation, and ensure that the device performance meets the specifications listed in the Technical Manual. In fact, you exchange potential device flexibility in exchange for the device to achieve the lowest instability, although your final circuit may have some design instability. When the performance of a programmable device is different from what you expect, is it because the device fails to meet the technical specifications, or is it because you are not using the device properly, these are difficult to determine.
For the tools and development systems provided by Programmable IC manufacturers, you should also use them smoothly, because you need these software and programming units in design. Similarly, this situation is different from most Analog design practices. The simulation design uses the technical specifications and device models provided by the manufacturer, but supports simulation tools such as spice, or EDA tools from various sources. In addition, such different design methods do not require programming systems.

Back to measure various options
If your applications can run at a lower performance level, or the requirements are flexible or difficult to define, programmable analog devices may be the right choice. You can change the parameters of a selected device. You can also change the performance of the circuit, or even change its topological layout in the process, reduce inventory and exclusive products on the material list, and develop adaptive design, using these devices allows the system processor to change the characteristic parameters of the analog signal processing link, you may also protect your design secrets to some extent.
These devices are unique, and you can only use them in a certain range of general projects. You also need dedicated design and simulation tools, so you lose the wide range of manufacturers and opportunities that most simulation designs will enjoy. At the same time, we also need to consider the support for applications, how much design can be used for the standard simulator, and the time required to complete the design.
If your design requires flexibility, you may feel the ability to program certain functions using standard unprogrammable devices and mechanically or electrically driven methods, it may be a good alternative.