problem to remove dc offset from different signals

[mehdi33]

Hi !

I work for a signal processing subject. I acquire three different signals from three accelerometers ( Position x,y and z). The sampling frequency that I chose is 25 Khz.

My problem is at the subtraction of the offset from the signals . the following error appears : "Error occured at subtract -1802 in READ-TDMS.vi

I really need help! Here is what i have done

Thanks!

[AdamTrojak · Accepted solution]

Is this what you ment?

[billko · Accepted solution]

Wouldn't it just be simpler to have AC coupling on each accelerometer?

[mehdi33]

Hi AdamTrojak !

Yes that's what I wanted. Thanks for your help! 

[mehdi33]

 Hi billko!

Yes you are right. I did not know at the beginning of my tests that AC coupling eliminates the offset of the signal.

Thanks for the information!

[AdamTrojak]

Hello! 

I'm glad that I could help.

If you find my post as solution - mark it as solution 😉

[billko]

---

@mehdi33 wrote:

 Hi @billko!

Yes you are right. I did not know at the beginning of my tests that AC coupling eliminates the offset of the signal.

Thanks for the information!

---

Hi, mehdi33:

An additional reason why I suggested AC coupling was, as a general rule, you should clean up your signal as much as possible with hardware before subjecting it to measurment instruments.  Let's take a situation applicable to your situation, but using numbers convenient for discussion.

Let's say you have a setup where your accelerometer amplifier accepts a peak signal of 5 Volts.  Your accelerometer puts out a signal that has a peak voltage of 3 Volts.  We're okay right now.  But what if the DC offset is +2.5 Volts?  Now you have a 3 Volt signal plus a 2.5 Volt DC offset = 5.5 Volts!  Assuming the manufacturer was exact and not conservative with the amplifier spec, you're probably going to get some clipping!  Maybe a worst case scenario is that you blow something up!  No amount of software compensation will allow you to recover the original waveform - or damaged equipment - (although there are some algorithms that can do a reasonable guesstimate of what the signal used to look like).

Just my two cents worth of input.  😄

[johnsold]

One caution: If you are using gravity to do some initialization or "calibration" of the accelerometer, then you will need DC coupling.

Lynn

[billko]

---

@johnsold wrote:

One caution: If you are using gravity to do some initialization or "calibration" of the accelerometer, then you will need DC coupling.

Lynn

---

Oh, I actually did not know that!

[mehdi33]

Hello!

Thank you very much for your advice. it helped me to better differentiate between the AC coupling and DC coupling!

Best regards!