TI中文支持网Part Number:LDC1101
Hi,
I want to detect the vibration characteristics of a 10mm-thick aluminum alloy crossbar with a measurement range of 0.5–3mm, using the LHR (High-Resolution Inductance) mode. The sampling rate is set to 10 kSPS.
Now I need to design a sensor. According to Optimizing L Measurement Resolution for the LDC161x and LDC1101.PDF, for the LDC161x, the highest effective resolution typically occurs when fSENSOR is between 400kHz and 6MHz (for fREF=40MHz). However, it does not specify the maximum useful resolution for the LDC1101. As stated in LDC1101 Data Sheet (9.1.11 Sensor Input Divider), the reference clock frequency should be greater than 4 times the sensor frequency for optimum measurement resolution: fCLKIN>4⋅fSENSOR−MAX.
Taking these factors into account, in order to optimizing L measurement resolution, within the allowable sensor frequency range (500 kHz to 10 MHz), should the maximum useful sensor frequency be designed as 4MHz when fCLKIN=16MHz?
Thanks.
Taylor：
Hi,
We have received your post and the investigation will take some time. Thank you for your patience.
,
Taylor：
Hi,
Many systems which use the LDC1101, the actual sensor RP, sensor frequency, or sensor inductance is not necessary to determine the target position. For your application setup your system to get the max resolution and sample rate that you require.
Make sure your sample rate is more than 8 times the vibration frequency (Nyquist rate) in order to see the actual vibration.
The sensor operating frequency can also influence the resolution. See section 2.4 of the app note below and also figure 4.
https://www.ti.com/lit/an/snoa944/snoa944.pdf
,
le liu：
Hi,
In Figure 4 of the following app note. https://www.ti.com/lit/an/snoa944/snoa944.pdf
If fREF = 16MHz, fSENSOR / fREF = 0.2 , RCOUNT = 0x1000, The Normalized LDC1101 Resolution = 6E-7 (Hz/Hz), which means the expected resolution is 16 MHz * 6E-7 = 11.2 Hz？However, under the same reference frequency and sensor frequency, why does increasing the RCOUNT result in a decrease in the expected resolution?
Thanks.
,
Taylor：
Hi,
I think you are misinterpreting figure 4. The different colors represent a different RCOUNT value. As RCOUNT is increased, normalized resolution(step size) is lower. This is consistent with all RCOUNT settings.
,
le liu：
Hello,
I would appreciate it if you could explain the relationship between the ratio of fSENSOR / fREF in Figure 4 and the L Measurement Resolution of LDC1101, as well as the trend. For example, when the ratio of fSENSOR / fREF is between 0.05 and 0.25, how does the LDC1101 Resolution change? I am quite puzzled by this.
Thanks.
,
Taylor：
There may not be any further information to explain this, please refer to the diagram and my explanation above
