Self-calibration type capacitance frequency conversion method
By employing a time-to-digital converter and a self-calibration method using a reference capacitor array, the temperature drift and stability issues in capacitance detection are resolved, resulting in improved accuracy and anti-interference capabilities, thus meeting the requirements for capacitance detection over a wide temperature range.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-31
- Publication Date
- 2026-03-10
AI Technical Summary
Existing capacitance detection technologies suffer from problems such as temperature drift, insufficient long-term stability, and weak anti-interference capabilities, failing to meet the requirements for wide temperature threshold, high precision, and maintenance-free detection.
A time-to-digital converter is used to replace the traditional ADC. Combined with an intelligently switched reference capacitor array and a dynamic temperature compensation algorithm, the capacitor frequency conversion is achieved through self-calibration. Calibration is performed using the time pulse width, and a second-order temperature compensation model is established.
It achieves an accuracy of ±0.5% across the entire temperature range of -40℃ to +125℃, improves its resistance to power supply ripple fluctuations to ±0.15%-5%, and reduces long-term drift to 0.5%/year, meeting the requirements for high accuracy and stability.
Smart Images
Figure CN121633629A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of electronic measurement of capacitance, and particularly relates to a self-calibration type capacitance frequency conversion method. BACKGROUND
[0002] The prior art has the following defects in the process of detecting capacitance: 1. Temperature drift problem The traditional capacitance voltage conversion circuit relies on an operational amplifier, and the input bias voltage of the operational amplifier drifts with temperature (typical value: ±1 muV / ℃), which causes the output error to exceed ±5% under a 100℃ temperature difference.
[0003] 2. Insufficient long-term stability The capacitance value decays due to aging of the reference capacitance medium, and a certain automotive electronics case shows that the detection error reaches +7.3% after one year of work (data source: SAE Paper 2023-01-1206).
[0004] 3. Weak anti-interference ability The power supply ripple is directly coupled to the output end, and 5% power supply fluctuation can cause ±1.2% reading deviation.
[0005] In summary, the existing capacitance detection technology cannot meet the detection requirements of wide temperature threshold, high precision, and maintenance-free, and therefore the application discloses a self-calibration type capacitance frequency conversion method. SUMMARY
[0006] The application discloses a self-calibration type capacitance frequency conversion method, which replaces the traditional ADC with a time-to-digital converter, combines an intelligent switching reference capacitance array and a dynamic temperature compensation algorithm, and realizes self-calibration in all working conditions.
[0007] The application is implemented through the following technical solutions: A self-calibration type capacitance frequency conversion method, the measured capacitance is connected to a charging loop through a switch matrix, the measured capacitance is charged, the charge of the measured capacitance is transferred to an integration capacitance, and an integration voltage is generated; the integration voltage is input into a five-stage ring oscillator to convert the integration voltage into a square wave signal, the square wave signal is amplified in pulse width by a time amplifier; a calibration engine is introduced, the calibration engine starts calibration by detecting calibration parameters, switches the reference capacitance array to the measurement channel in sequence, obtains the current pulse width, and compares it with the factory calibration value of the reference capacitance; if the comparison deviation exceeds the threshold, a proportional factor is calculated, and the proportional factor is superimposed with a second-order temperature compensation model, and finally the calibrated measured capacitance value is output.
[0008] In order to better implement the application, further comprising the following steps: Step 1, real-time judge whether calibration is needed based on temperature change rate, continuous working time, and power-on position mark, if calibration is needed, then go to step 2, if calibration is not needed, then stop; Step 2, switch the measurement channel from the measured capacitance to the reference capacitance, and obtain the reference measurement value; Step 3, measure the pulse width of the reference capacitance, and obtain the original measurement value of the reference capacitance under the current working environment; Step 4, calculate the calibration coefficient based on the deviation value between the original measurement value of the reference capacitance and the factory calibration value; Step 5, measure the to-be-measured capacitance, obtain the measured value of the to-be-measured capacitance before calibration, and update the measured value of the to-be-measured capacitance based on the calibration coefficient.
[0009] In order to better realize the present application, further, the step 1 specifically comprises: using a temperature sensor with an accuracy of ±0.1℃, performing temperature sampling every 10ms, and calculating the temperature change rate, when the temperature change rate is greater than 2℃ / min, triggering calibration; recording the continuous working time of the capacitance, if the continuous working time is greater than or equal to 24h, triggering calibration; detecting the power-on position mark, if the power-on position mark is detected, triggering calibration.
[0010] In order to better realize the present application, further, in the step 2, a reference capacitance array is established, an estimated value of the to-be-measured capacitance is estimated, and the reference capacitance closest to the estimated value is automatically matched.
[0011] In order to better realize the present application, further, the original measurement values of each reference capacitance in the reference capacitance array are measured in sequence, the measured ratios between the reference capacitances are calculated, and the measured ratios are compared with the calibration ratios, if the deviation between the measured ratios and the calibration ratios is greater than or equal to 15%, the current reference capacitance array is reported as an error.
[0012] In order to better realize the present application, further, in the step 3, the reference capacitance is charged to 1.8V, so that the charge of the reference capacitance is transferred to the integration capacitance, and the integration voltage is generated; then the integration voltage is input into a five-stage ring oscillator, the integration voltage is converted into a square wave signal through the five-stage ring oscillator; the square wave signal is amplified by eight times through a time amplifier to obtain the pulse width, as the original measurement value of the reference capacitance under the current working environment.
[0013] In order to better realize the present application, further, in the step 4, if |T_meas-T_ref|>50+0.1T_ref, the calibration coefficient is calculated; wherein: T_meas represents the original measurement value; T_ref represents the factory calibration value.
[0014] In order to better realize the present application, further, the calculation formula of the calibration coefficient is: ; wherein: represents a calibration coefficient; K represents a preliminary estimated calibration coefficient; k1 represents a first-order temperature drift coefficient; k2 represents a second-order temperature drift coefficient; k3 represents a capacitance aging attenuation coefficient; represents a temperature change value.
[0015] Compared with the prior art, the present application has the following advantages and beneficial effects: (1) The present application uses a time-to-digital converter to replace a traditional ADC, converts the capacitance charge into a time pulse width, and avoids operational amplifier temperature drift; (2) The present application configures a multi-capacitance silicon nitride thin film reference capacitor array, and triggers self-calibration through a temperature change rate > 2℃ / min or a timing event; (3) The present application establishes a compensation model containing a second-order temperature term and an aging term, the coefficients are solidified through three-temperature calibration, and a ±0.5% precision is achieved in a full temperature range of -40℃ to +125℃, the anti-power ripple capability is improved to ±0.15%-5% fluctuation, and the long-term drift is reduced to 0.5% / year. BRIEF DESCRIPTION OF DRAWINGS
[0016] Figure 1 is a system circuit schematic diagram of the self-calibration type capacitance frequency conversion method; Figure 2 is a TDC circuit schematic diagram; Figure 3 is a charge transfer timing diagram; Figure 4 is a dynamic calibration flowchart. DETAILED DESCRIPTION
[0017] Embodiment 1: A self-calibration type capacitance frequency conversion method of the present embodiment, the to-be-measured capacitance is connected to a charging loop through a switch matrix, the to-be-measured capacitance is charged, the charge of the to-be-measured capacitance is transferred to an integration capacitor, and an integration voltage is generated; the integration voltage is input into a five-stage ring oscillator to convert the integration voltage into a square wave signal, the square wave signal is amplified in pulse width by a time amplifier; a calibration engine is introduced, the calibration engine starts calibration by detecting calibration parameters, acquires the current pulse width by sequentially switching the reference capacitor array to the measurement channel, and compares it with the factory calibration value of the stored reference capacitor; if the comparison deviation exceeds the threshold, the proportional factor is calculated, and the proportional factor is superimposed with a second-order temperature compensation model, and finally the calibrated to-be-measured capacitance value is output.
[0018] Further, the following steps are included: Step 1, real-time judge whether calibration is needed based on temperature change rate, continuous working time, and power-on position mark, if calibration is needed, go to step 2, if calibration is not needed, stop; Step 2, switch the measurement channel from the measured capacitance to the reference capacitance, and obtain the reference measurement value; Step 3, measure the pulse width of the reference capacitance, and obtain the original measurement value of the reference capacitance under the current working environment; Step 4, calculate the calibration coefficient based on the deviation between the original measurement value of the reference capacitance and the factory calibration value; Step 5, measure the to-be-measured capacitance, obtain the measured value of the to-be-measured capacitance before calibration, and update the measured value of the to-be-measured capacitance based on the calibration coefficient.
[0019] Embodiment 2: The embodiment discloses a self-calibration type capacitance frequency conversion method, which is optimized on the basis of embodiment 1. The step 1 specifically comprises: using a temperature sensor with an accuracy of ±0.1℃ to perform temperature sampling every 10ms, and calculating a temperature change rate; when the temperature change rate is greater than 2℃ / min, calibration is triggered; recording the continuous working time of the capacitance; if the continuous working time is greater than or equal to 24h, calibration is triggered; and detecting a power-on position mark; if the power-on position mark is detected, calibration is triggered.
[0020] In the step 2, a reference capacitance array is established, an estimated value of the to-be-measured capacitance is estimated, and the reference capacitance closest to the estimated value is automatically matched.
[0021] The original measurement values of each reference capacitance in the reference capacitance array are measured in sequence, the measured ratios between the reference capacitances are calculated, and the measured ratios are compared with the calibration ratios; if the deviation between the measured ratios and the calibration ratios is greater than or equal to 15%, an error is reported for the current reference capacitance array.
[0022] In the step 3, the reference capacitance is charged to 1.8V, so that the charge of the reference capacitance is transferred to an integration capacitance, and an integration voltage is generated; then the integration voltage is input into a five-stage ring oscillator, the integration voltage is converted into a square wave signal through the five-stage ring oscillator; after the square wave signal is amplified by eight times through a time amplifier, the pulse width is obtained as the original measurement value of the reference capacitance under the current working environment.
[0023] In the step 4, if |T_meas-T_ref|>50+0.1T_ref, the calibration coefficient is calculated; wherein: T_meas represents the original measurement value; and T_ref represents the factory calibration value.
[0024] The calculation formula of the calibration coefficient is: ; wherein: represents a calibration coefficient; K represents a preliminary estimated calibration coefficient; k1 represents a first-order temperature drift coefficient; k2 represents a second-order temperature drift coefficient; k3 represents a capacitance aging attenuation coefficient; represents a temperature change value.
[0025] The remaining part of the embodiment is the same as that of Embodiment 1, and thus will not be described again.
[0026] Embodiment 3: This embodiment discloses a self-calibration capacitance frequency conversion method, which is optimized on the basis of Embodiment 1 or 2, as shown in the formula (1). Figure 1 As shown in the formula (1), it is a system circuit schematic diagram of the self-calibration capacitance frequency conversion method based on time-to-digital conversion and a reference capacitance array according to the present application, which describes a complete signal chain of capacitance voltage conversion. The measured capacitance and the reference capacitance array Cr1-Cr4 are switched and accessed through a switch matrix.
[0027] Charging phase: close SW1A, charge to the reference voltage V ref =1.8V, charging time T chg =2μs, then the charge quantity is Q x =C x ×V ref =C x ×1.8.
[0028] Charge transfer, convert the charge Q x into a measurable voltage signal: open SW1A, wait for a 50ns dead time, and close SW1B. C x Charge is transferred to the integration capacitance C int (2nF), to generate an integration voltage V int . The transfer time is strictly controlled at 5μs by a TCXO frequency division clock.
[0029] ; Time conversion: V int A square wave signal with a frequency proportional to voltage is generated by driving a five-stage ring oscillator, and a single-stage delay time t d is set: ; wherein R n =1.2KΩ, which is the NMOS on-resistance; R p =2.8KΩ, which is the PMOS on-resistance; C load is the load capacitance, C load =0.1pF; and t d =138ps is calculated.
[0030] Output square wave frequency: ; Where: f RO represents the output square wave frequency; f0 represents the free oscillation frequency of the five-stage ring oscillator when the input voltage is 0; k represents the modulation sensitivity of the input voltage to the oscillation frequency.
[0031] ; Where: N stage represents the number of stages of the inverter.
[0032] Time amplification: expand the pulse width to improve the counting resolution. The time amplifier (TA) adopts a capacitor charge redistribution structure, the input capacitor C1=10fF, the output capacitor C2=80fF, and the gain formula: ; Then the output pulse width is: ; Where: ; T out represents the output pulse width; T in represents the input pulse width; the square wave signal is amplified in pulse width by the time amplifier, and the square wave period is amplified by 8 times.
[0033] Pulse counting: the STM32 converts the time signal into a digital quantity N: ; Where: f clk represents the counter clock frequency, which is 10GHz.
[0034] Calculate Cx raw : According to the digital quantity N, the measured capacitance value is deduced: ; Let S=f clk ×C int ×G, D=f0×C int , then , the Cx raw is: ; Calibration output: the calibration engine combines temperature sensor data and EEPROM calibration values to execute the compensation formula: ; Where: represents the calibration coefficient; K represents the preliminary estimated calibration coefficient; k1 represents the first-order temperature drift coefficient; k2 represents the second-order temperature drift coefficient; k3 represents the capacitance aging attenuation coefficient; represents the temperature change value.
[0035] As Figure 2 shown, the TDC circuit in the application is a schematic diagram, which consists of a ring oscillator (RO) and a time amplifier (TA). The ring oscillator consists of a closed loop of five CMOS inverters, and the single-stage delay t d = 0.69 x (R n + R p ) x C load .
[0036] Wherein: R n = 1.2KΩ; R p = 2.8KΩ; C load = 0.1pF, t d = 138ps is obtained by calculation.
[0037] The input V int controls the power supply voltage of the inverter, realizing voltage-frequency conversion. The time amplifier adopts the principle of capacitor charge redistribution: the rising edge of the input pulse closes S1, charging C1 to VDD; S2 is closed during the high level, and C1 discharges to C2. The output pulse width: ; As Figure 3 shown, the charge transfer timing diagram of the application describes the timing control of the charge transfer process. As Figure 4 shown, the dynamic calibration flowchart of the application is the decision logic and algorithm process of STM32. First, monitor the trigger condition: the temperature change rate monitored by the temperature sensor > 2℃ / min, or power-on event or 24-hour timing reaches. Second, traverse the reference capacitor: switch the switch matrix to connect the reference capacitor Cr1~Cr4 in turn, for each reference capacitor, get the current pulse width, read the factory calibration value, threshold = 50 clock cycles, calculate the calibration coefficient , . Finally, the calibration coefficient is applied: real-time compensation is made for the Cri measurement value: ; Superimposed temperature / aging correction: .
[0038] Fault-tolerant mechanism: if the calibration fails for 3 times in a row, trigger an error code and switch to the backup capacitor.
[0039] The rest of this embodiment is the same as embodiment 1 or 2, so it will not be described here.
[0040] The above is only the preferred embodiment of the application, not any form of limitation on the application, any simple modification or equivalent change according to the technical essence of the application to the above embodiment, falls within the protection scope of the application.
Claims
1. A self-calibrating capacitance-to-frequency method, comprising: The to-be-tested capacitor is connected to a charging circuit through a switch matrix, and the to-be-tested capacitor is charged to make the charge of the to-be-tested capacitor transferred to an integration capacitor to generate an integration voltage; the integration voltage is input into a five-stage ring oscillator to convert the integration voltage into a square wave signal, and the square wave signal is subjected to pulse width amplification by a time amplifier; A calibration engine is introduced, which detects calibration parameters to start calibration, switches the reference capacitor array to the measurement channel in sequence, obtains the current pulse width, and compares it with the stored factory calibration value of the reference capacitor; if the comparison deviation exceeds the threshold, the scaling factor is calculated, and the scaling factor is superimposed with a second-order temperature compensation model, and finally the calibrated to-be-tested capacitor value is output.
2. A self-calibrating capacitance-to-frequency method as claimed in claim 1, characterized in that, The method comprises the following steps: Step 1, based on the temperature change rate, the continuous working time, and the power-on position flag, it is judged in real time whether calibration is needed, if calibration is needed, step 2 is entered, if calibration is not needed, it is stopped; Step 2, the measurement channel is switched from the measured capacitor to the reference capacitor to obtain the reference measurement value; Step 3, the pulse width of the reference capacitor is measured to obtain the original measurement value of the reference capacitor under the current working environment; Step 4, the calibration coefficient is calculated based on the deviation between the original measurement value of the reference capacitor and the factory calibration value; Step 5, the to-be-tested capacitor is measured to obtain the measured value of the to-be-tested capacitor before calibration, and the measured value of the to-be-tested capacitor is calibrated and updated based on the calibration coefficient.
3. A self-calibrating capacitance-to-frequency method according to claim 2, wherein, The step 1 specifically comprises: using a temperature sensor with an accuracy of ±0.1℃, performing temperature sampling every 10ms, and calculating the temperature change rate, when the temperature change rate is greater than 2℃ / min, the calibration is triggered; record the continuous working time of the capacitor, if the continuous working time is greater than or equal to 24h, the calibration is triggered; detect the power-on position flag, if the power-on position flag is detected, the calibration is triggered.
4. A self-calibrating capacitance-to-frequency method according to claim 3, wherein, In step 2, a reference capacitor array is established, an estimated value of the to-be-tested capacitor is estimated, and the reference capacitor closest to the estimated value is automatically matched.
5. A self-calibrating capacitance-to-frequency method according to claim 4, wherein, The original measurement values of each reference capacitor in the reference capacitor array are measured in sequence, the measured ratio between the reference capacitors is calculated, and the measured ratio is compared with the calibrated ratio, if the deviation between the measured ratio and the calibrated ratio is greater than or equal to 15%, the current reference capacitor array is reported as an error.
6. A self-calibrating capacitance-to-frequency method according to claim 5, wherein, In step 3, the reference capacitor is charged to 1.8V, so that the charge of the reference capacitor is transferred to the integration capacitor, and the integration voltage is generated; then the integration voltage is input into a five-stage ring oscillator to convert the integration voltage into a square wave signal; the square wave signal is amplified by eight times by a time amplifier to obtain the pulse width as the original measurement value of the reference capacitor under the current working environment.
7. A self-calibrating capacitance-to-frequency method according to claim 6, wherein, In step 4, if |T_meas-T_ref|>50+0.1T_ref, the calibration coefficient is calculated; Wherein: T_meas represents the original measurement value; T_ref represents the factory calibration value.
8. A self-calibrating capacitance-to-frequency method according to claim 7, wherein, The calculation formula of the calibration coefficient is: ； wherein: represents a calibration coefficient; K represents a preliminary estimated calibration coefficient; k1 represents a first-order temperature drift coefficient; k2 represents a second-order temperature drift coefficient; k3 represents a capacitance aging attenuation coefficient; represents a temperature change value.