1986 views|0 replies

3836

Posts

19

Resources
The OP
 

About using a single chip microcomputer to read external voltage ADC impedance matching [Copy link]

This post was last edited by fish001 on 2020-9-13 16:40

The reference voltage of the microcontroller is generally 3.3V. If the external signal exceeds the AD measurement range, the resistor voltage division method can be used, but attention should be paid to the impedance matching problem. For example, the analog-to-digital input impedance of SMT32 is about 10K. If the external voltage divider resistor cannot be much smaller than this resistance value, the signal source output impedance will be large and the AD input impedance will be small, so the input impedance will divide the voltage of the signal source signal, which will eventually lead to a large voltage reading error.

  Therefore, when using a single-chip microcomputer to read external signal voltage, the external voltage divider resistor must be a smaller resistor, or if there is a requirement for power consumption, a large resistance voltage divider can be selected, and then a voltage follower can be used for impedance matching (the input impedance of the voltage follower can reach several megohms, and the output impedance is several ohms or even smaller). If the output impedance of the signal source is large, a voltage follower can be used for matching and then a resistor voltage divider can be connected.

When selecting an external ADC chip, pay attention to its type (SAR, switched capacitor, FLASH, dual integration, Sigma-Delta). Different types of ADC chips have different input impedances.

1. SAR type: This type of ADC has a large internal resistance, usually above 500K. Even for ADCs with low impedance, the impedance is fixed. Therefore, as long as the internal resistance of the source being measured is stable, it is equivalent to a resistor voltage divider and can be calibrated;

2. Switched capacitor type: such as TLC2543, which requires very low input impedance to quickly charge the internal sampling capacitor. At this time, it is best to have a low resistance source, otherwise it will cause errors. If it is really not possible, you can connect a large capacitor in parallel externally. After each sampling, the voltage of the large capacitor does not drop much. Therefore, after connecting an external large capacitor in parallel, the switched capacitor input can be equivalent to a pure resistive impedance and can be corrected;

3. FLASH type (direct comparison type): Most high-speed ADCs are direct comparison type, also known as flash type (FLASH), which are generally low impedance. They require a low resistance source. They are purely resistive to the outside and can be directly connected to the op amp;

4. Double integral type: Most of this type have extremely high input impedance, and there is almost no need to consider impedance issues;

5. Sigma-Delta type: This is the most accurate ADC type currently. The following issues need to be paid special attention to:

a. Measurement range problem: SigmaDelta ADC belongs to switched capacitor input and must have a low resistance source. Therefore, in order to simplify the external design, most of them have integrated buffers. When the buffer is turned on, it presents high resistance to the outside and is easy to use. But be careful, the buffer is actually an op amp. Then there must be restrictions on the upper and lower rails. Most buffers have a lower rail of 50mV and an upper rail of AVCC-1.5V. In this application, the common mode input range is greatly reduced, and it cannot measure 0V. Be especially careful! It is generally used in bridge measurements because the common mode range is around 1/2VCC. Don't worry too much about the zero vote of the buffer, it is easy to correct through the internal zero register;

b. The problem of RC filter at the input: SigmaDelta ADC belongs to switched capacitor input and works well on low impedance sources. But sometimes, in order to suppress common mode or suppress signals outside the Nyquist frequency, it is necessary to add RC filter at the input. Generally, the DATASHEET will give a table of the maximum allowable input impedance and the relationship between C and Gain. A very strange characteristic at this time is that the larger C is, the smaller the maximum input impedance must be! Many people may be puzzled at first, but in fact, it is easy to understand as long as you think about the charging characteristics of the capacitor. Another compromise is to make C very large, much larger than the sampling capacitor Cs (usually 4~20PF) by millions of times, then the input is equivalent to pure resistance, and the voltage division error can be corrected with the GainOffset register.

c. The op amp must not be directly connected to the SigmaDelta ADC! As mentioned earlier, the switched capacitor input circuit uses a sampling capacitor to sample from the input end. Each time it is connected in parallel with the op amp, it will present a low impedance, which will be divided by the op amp output impedance, causing the voltage to drop. The negative feedback will start to correct immediately, but the op amp slew rate (SlewRate) is limited and cannot respond immediately. This will cause an instantaneous voltage drop. When the sampling is almost completed, it is equivalent to high impedance. The op amp output voltage rises, but the slew rate makes it too late for the op amp to correct, resulting in overshoot. This is the most critical moment of sampling end. Therefore, the op amp must be connected to the SD ADC through a resistor and capacitor (connected as a low pass). The RC relationship must obey the rules described in the datasheet.

d. Differential input and bipolar issues: SD ADCs can all have differential inputs and support bipolar inputs. However, the bipolarity here does not mean that negative voltage can be measured, but the voltage between the two pins Vi+ and Vi-. Assuming that Vi- is connected to AGND, the negative voltage measurement range will not exceed -0.3V. The correct connection method is that the common mode of Vi+ and Vi- is differentially input between -0.3 and VCC. A typical example is a bridge. Another example is that Vi- is connected to Vref, and the voltage of Vi+ to Vi- allows bipolar input.

This post is from Microcontroller MCU
 

Guess Your Favourite
Just looking around
Find a datasheet?

EEWorld Datasheet Technical Support

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京B2-20211791 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号
快速回复 返回顶部 Return list