02-01-2010 11:21 PM - edited 02-01-2010 11:24 PM
Hi all,
I am working on one DAQ system. Sensor intend to use is ICP accelerometer 350B04. DAQ card 6052E. Signal conditioning card, SCXI 1530 with SCXI Chassis-1001.
The ICP sensor is 350B04, which needs 4mA current excitaion and is connected to a BNC connector through coaxial cable.
http://pcb.com/spec_sheet.asp?model=350B04
According to SCXI 1530 user manual, 4mA 24V compliant current source was provided internally. I know in order to use this internal current excitation, I need to set it as "internal excitation" and can program 4mA as current source.
I have couple of question on the connections.
1. Can I directly connect the ICP sensor to SCXI 1530 through BNC interface provided by SCXI 1530? My assumption is, after connected, scxi 1530 will provide 4mA current source to ICP sensor. Then the sensor is excited and give an output AC signal with 8-14 DC bias vlotage, which is actually the input signal of SCXI 1530 in my case.
Since the amplifier of SCXI 1530 is AC coupled. I guess the DC voltage will be removed and left AC signal for further operation?
2.
Could any one of you explain to me how could this happen througn BNC connector and coaxial cable? I mean how to send the exctiation current and get the excited signal back in the mean time through coaxial cable? Can I somehow measure this DC bias vlotage to make sure my accelerometer is working fine?
Thank you very much!
02-02-2010 02:39 PM
Hi oly,
You are correct with your understand of the IEPE signals and the sensor signals. The IEPE excitation and the sensor signals will be present on the lines in the coax cable. The excitation will be a DC signal while the sensor signal will be AC. The AC coupled input to the 1530 means that the data acquired will be the sensor signal because the DC signal from the excitation is being blocked. The ICP sensor and the 1530 can be directly cabled together. It could be possible to measure the DC offset on the coax by using a bnc splitter with a high impedence measurement device. This could affect your actual measurement, so it would not be recommended for anything other than verifying the DC voltage on the line.
02-02-2010 04:46 PM - edited 02-02-2010 04:51 PM
Hi Steve_B,
Thank you for the clear explaination!
The accelerometer seems working fine now. We can see amplitude of the output is increasing when the accleration increases.
I got one more question. Based my observations, the readout data from the acceleromter are the same whether I turn on the current excitation or not. I am using traditional DAQ and the AI parameter VI to get the current excitions we are using. The output did not change if the current is 4mA or 0. I am pretty sure I did turn the current excitaion on and off. Any idea?
One more question, is the default setting for SCXI 1530 low pass filter off?
02-02-2010 05:33 PM - edited 02-02-2010 05:38 PM
One more quesiton, from MAX, go scxi 1530 properties, there is one setting called ground referencing or something. What is the difference turning this on or off? Thank you!
If the accelerometer is grounded to earth, I should turn off this ground reference? What this really change inside the SCXI card?
02-03-2010 04:51 PM
Hi oly,
I'm happy to hear you have been able to get your accelerometer working. It sounds like you are using the right VI to control the excitation current, however it sounds like the current was not disabled. Can you provide a screenshot of your VIs? Do you still measure a DC offset when you turn off the excitation current?
The Low Pass filter can be set to 2.5k, 5k, 10k, and 20kHz and this can be configured in MAX. It should default to one of these frequencies and you can change this setting following the steps listed in the User Manual on page 19.
The ground referencing property will determine whether or not we introduce a reference to ground for the measurement before our amplifier which will restrict the drift that can happen when connecting to a floating sensor. If your sensor is ground referenced you will want to turn off the ground referencing to avoid a potential ground loop. There is a description of this on page 29 of the User Manual.
SCXI-1530/1531 User Manual and Specifications
http://digital.ni.com/manuals.nsf/websearch/6F3ACDC99AAD6153862568E30063CB5B
02-03-2010 06:24 PM
Hi Steve,
Thank you for your insightful words!!!
I do not have a BNC splitter on hand. So I could not measure the DC offet. I am using example VI provided by NI to test SCXI 1530 now,which should be in the folder examples/daq/scxi/scxi 1530 if I did not remember wrong.
I am not using program to control current excitation. I am using the MAX and set current excitaion to 4mA or off respectively, however, I got the same output.
The reason I was asking questions about low pass filter is that I had gone through the user manual and found on page 2-5 "The minimum scan rate for the module must be at least 10 times the highest filter bandwidth setting. "
Say my single channel signal is 100Hz. I have to user 2.5KHz low pass filter if the filter can not be turned off. Do I still need 25 KHz sample rate?
02-08-2010 10:10 AM
Hi oly,
I'm sorry for the delay in my response.
It would be good to test the DC offset if you are able to get a bnc splitter in order to verify the excitation is turning on and off. You could potentially use a resistor as a load for the 4mA excitation and measure the voltage drop across the resistor when the excitation is turned on and off. You would want to use a load that is small enough not to cause an overvoltage or an issue with the power dissipation.
The manual does state that a minimum value of 10x greater than the filter bandwidth should be used as the scan rate, this is somewhat of a recommendation. The minimum oversample that is necessary to return valid frequency information is 2x the bandwidth based on Nyquist. In order to get valid time information a minimum oversample of 10x is typically what we recommend. Based on your description you could probably sample at 5kHz and be ok. This would satisfy the 2x Nyquist requirement which means that you should not be experiencing aliasing and you will have 50x oversample of your signal of interest.