If you have any questions or comments about this circuit please post them below.
It is mentioned that CN0288 is made for inductive LVDT. I want to apply this circuit for a capacitive sensor. Is it possible to use this circuit with capacitive sensor for measuring a displacement? I attach the scheme of capacitive sensor as an image. Sensor looks like two cylindrical variable capacitors which can be used in bridge.
Are you planning on using the AD598? I'm not sure if this will work. What is the sensor you are planning on using?
As far as I understand AD598 is a part of CN0288 and it has an analog signal at the out. For me it is better to have logic out. But if AD598 will suit for my sensor it will be OK.
Sensor that I want to apply is hand-made device and it looks like inductive LVDT, but it is capacitive. Value of capacitance is approximately 4 pF. It consist of two coaxial cylinders. Internal cylinder can move back and forth so the capacitance changes.
The AD598 drives an AC voltage on the Excitation Pins, if you need a DC signal it might not work.
I would like to use the CN0288 but I need a high pass filter for the LVDT secondary coils to remove the primary signals. Are there details for the AD598 A/B filtering. The LVDT is used for detection more than position.
Do you have any suggestions?
I have problem getting the bandwidth I need out of the demo board. We are looking at integrating the position sensing into a system rather than having a purchased LVDT signal conditioner so the CN-0288 looked like a good start. I calibrated the output using gage blocks and I have a secondary position sensor (Keyence laser position sensor) At DC and up to as fast as I can jerk the core in the system they track perfect but as I start to excite the system I loose output very quickly and at 120 Hz I get a 0.00035" output where my true motion is 0.005"
I also hooked up a very small LVDT (Measurement specialty MHR025) and put a stainless screw through it and two nuts as travel stops allowing me to flick it with a finger and the output wave form show a rise time of about 30-40ms. I tried to change the three caps that set the bandwidth to 0.1uF but that did not make a difference. I also attached a lead after the AD8615 thinking my DAQ system or the scope loaded the AD598 but that signal is identical to the one at the JC6 SMB output.
I'm at a loss here. Any advice is greatly appreciated.
CN0288 is designed specifically to work with an E100 LVDT. If the LVDT that you are using does not have the exact same operating characteristics (Voltage Transformation Ratio (VTR), target excitation frequency, etc.), then the passive components that set the operating characteristics of the AD598 will need to be recalculated to configure the system for your specific transducer.
The A and B inputs can never see a signal greater than 3.5 Vrms. If your LVDT has a more sensitive VTR than an E100 LVDT, then you may be saturating the inputs.
Also, C11 and C12 of CN0288 should be DNI. We're in the process of updating the BOM.
In the application I use a HR050 and it had quite similar characteristic. it has 290mV/V output, 2.5kHz, Slightly higher input and output impedance. The scaling in v/inch would differ but I cant see anything in the datasheet that is pointing on how the bandwidth could be reduced to ~25Hz vs 250Hz. As far as I can read the C2-C3 and C4 are the primary for setting the bandwidth and the other components deals mainly with the gains.
Am I missing something?
Should I remove C11 and C12?
Yes, please remove C11 and C12. These are incorrectly populated and should be DNI. However, these components have no association to system bandwidth.
You are correct that data sheet capacitors C2,C3 and C4 are the bandwidth capacitors. For a 250 Hz BW, they should be set to 400 nF, so 100 nF should result in a bandwidth even greater than 250 Hz. CN0288 also features a shunt capacitor across the output scaling resistor (C8 and C10) that will influence output settling time. There is also a low-pass RC filter between the AD598 and the AD7992 that will limit system bandwidth. These may require adjustments to achieve your bandwidth of interest.
What is the VTR of the HR050 in V/V/inch?
The sensitivity is about 2X. It is 5.8V/V/in since it is a 0.050 stoke unit. Input impedance is 430Ohm vs 660 on the E100 and output impedance is 4000Ohm vs 960Ohm for the E100. Its hard for me to see how any of the LVDT parameters would change the rise time/bandwith that much. I took the little MHR025 I used to measure the rise time and hooked it to a Measurement Specialties LVM-110 conditioner as we currently are using and the typical rise time I see when flicking the core about 0.010" is ~1ms on that card while I get about 25-30ms on the CN0288. I also have a CN0301 and that show the same slow response. I also get the exact same rise time at the J6 port as I get at the R14-C31 junction leading into the A/D. I will pull the C8/C10 and let you know what I find.
Thanks again! I reapply appreciate it!
You nailed it Ryan, The C10 is gone and now I am at 1ms. The ripple is up to 75mV but I will readjust the BW caps and see what I get.
After going back to original caps on the BW pins i'm down to 40mV ripple but after the R14-C31its only 12mV. Its mainly composed of a low level 400kHz signal and a 2.5kHz and 5 kHz so it should be easy to get rid of.
Are there by any chance a set of VI's available for CN0288 or CN0301?
I'm happy to have helped and glad to hear you had success! I had asked for the VTR parameters so that I could independently perform the passive component calculations. I wanted to ensure that there were no clipping issues in the system that manifested as a bandwidth issue.
The purpose of C8 and C10 is to provide additional ripple filtering, but as you found out, it comes at a trade off of reducing system bandwidth.
Are you looking for LabView VIs for processing the data from the AD7992? I would need to speak with the group who developed the board and the GUI to get the files used to make the CN0288 executable.
Please let me know if this is what you are looking for,
I’m very grateful for you helping on this and I think it will be of great help for others using the boards. It was very easy to find this thread so now it is available for others in the future. Maybe you want to post something in the CN0301 board in regards to this since I think that has the same issue.
Yes. I would like to be able to pull data from the AD7992 trough the Blackfin board using USB/labview if possible. It looks like the CN0288 app is written in Labview. Where I work IT even need a legal review of the license before I can install any .exe file. A VI I can load up and use with no IT involvement.
I've loaded the source files up onto our FTP site for you to use. ftp://ftp.analog.com/pub/cftl/CN0288/1.0.2/
Please note that by downloading the source files, we at ADI do not support software development or support for that. These files are entirely to assist you in any learnings that you may get from them. Hope this helps Ulf.
Hello Engineering Zone support community,
I am looking at viability of using the CN0288 LVDT Conditioning Circuit containing the AD598 integrated circuit to generate an excitation voltage of 7.07Vrms amplitude at a frequency of 1800Hz to the primary coil of an LVDT with which I am looking to supply the excitation to and signal condition the read out from the secondary coils. Having taken a brief look through the design procedure in the AD598 datasheet on page 6, I have determined that C1 (Capacitor between pins 6 & 7) sets the frequency of the excitation whereas R1 (resistor between pins 4 & 5) sets the amplitude of the excitation is that correct? I understand that more than just those two components must be changed in order to "tune" the circuit to match a different LVDT, however my concern and question is how significant is the excitation signal in terms of affecting the read out from the secondary coils if the circuit uses the difference and sum of these output voltages to perform it's ratiometric decoding and final output voltage?
AD598 relies upon an extremely linear transducer. The sum of the LVDT secondary amplitudes must be constant throughout the transfer function.However, the exact amplitudes and frequency within data sheet specifications should not influence the circuit performance. The circuit is designed ratio-metric to account for deviation in amplitude/ frequency. We do recommend maximizing the voltage on the secondary outputs to improve system dynamic range.
Retrieving data ...