amplifier and filter

[Gram]

Many thanks, Logan Kunitz, I have tried the accelerometer and the charge amplifier together; however, I cannot get a proper signal. The accelerometer is with built-in electronics (different technics?) and does not work with the charge amplifier? I am not clear about that.

[Buechsenschuetz]

A digital filter cannot replace a hardware filter on the input completely, but in some cases it will be useful. For example, it will eliminate spurious 'digital noise' generated by the ADC. So I think in your application it will be useful to set the digital low pass filter to a value close to the bandwidth of the sensor.

[malef]

ISOTRON is the trademark of ENDEVCO accelerometers requirering a signal conditioner with constant current power excitation. This technique is commonly called IEPE. See more datails at: http://digital.ni.com/public.nsf/websearch/e96dccbd929ef9e886256c93006d9ec5
Your amplifier Type 2635 is a charge conditioner and cant be used with IEPE sensors. All major suppliers of vibration equipment can recommend suitable IEPE signal conditioners.

Greetings

Manfred

[Gram]

Thanks, Buechsenschuetz. Your advice is very helpful to me.  I have tried  different low pass filter values, and found that the filtered signal, especially filtered with the lowest value, had reduced magnitudes, however, the major frequency was nearly the same. Is this right?

Manfred, thanks for your explaination and excellent link.

[Logan_K]

In light of Manfred's response I didn't look closely at the accelerometer so I totally overlooked the fact that it is IEPE compatible. National Instruments makes several devices that can be used for measuring IEPE type accelerometers depending on what type of platform you prefer. Almost all of our accelerometer modules provide appropriate anti-aliasing filters which will be perfect for your application, meaning that you won't need any additional hardware. We have a SCXI-1530 module, PCI-4474, and a NI USB-9233. I have used our devices with several different types of Endevco accelerometers, so I know that they should work well for you.

Regards,

Logan

[Gram]

Many thanks, Logan.

[Buechsenschuetz]

In most cases filter cut-off frequencies are indicated for the -3db point, i.e. signals with the cut-off frequency will be reduced significantly. To make sure you should set the cut-off frequency at 1.5x or 2x the upper frequency limit of your signal source.

Readings of filtered signals also depend on the waveform of the input signal. Any signal different from sinusoidal form contains harmonics of 2f, 3f etc. Depending on the cut-off frequency these harmonics are cut off more or less, and readings will be different from actual signal amplitude when approaching the cut off frequency. For example, if you want to measure repetitive spikes with a given frequency, depending on the rise time of the spikes you will get readings which have little to do with actual signal amplitude.

[Buechsenschuetz]

In most cases filter cut-off frequencies are indicated for the -3db point, i.e. signals with the cut-off frequency will be reduced significantly. To make sure you should set the cut-off frequency at 1.5x or 2x the upper frequency limit of your signal source.

Readings of filtered signals also depend on the waveform of the input signal. Any signal different from sinusoidal form contains harmonics of 2f, 3f etc. Depending on the cut-off frequency these harmonics are cut off more or less, and readings will be different from actual signal amplitude when approaching the cut off frequency. For example, if you want to measure repetitive spikes with a given frequency, depending on the rise time of the spikes you will get readings which have little to do with actual signal amplitude.

[Gram]

Buechsenschuetz, thanks kindly for all your explainations.

[Kay Kay]

I'm wondering if these signal conditioning boards provided by NI are also useful in low power measurements. I've studied the signal conditioning tutorial of NI and realized that the signals (e.g. high voltage, noisy environment, extreme high or extreme low) which are difficult to be measured using DAQ boards could be measured using the signal conditioning circuits. I’m also aware of the fact that signal conditioning maximizes the accuracy. So, I’m interested to know if these signal conditioning circuits might be helpful in reducing the noise and for more accurate measurements of the low power in the range 1-30 milli Amperes. I still am not sure of how they might be helpful in low power measurements. I appreciate the ideas. Also, if you have prior experience working on low power measurements, please share your views.