How to smooth the signal?

[RTSLVU · Accepted solution]

Also consider a "running average"...

Say you are taking a measurement every 100mS

Average the reading over 1 second (10 readings) and save or display the average.

Most power analyzers do this and offer several levels of averaging depths. One of my Yokogawa power analyzers will average from 2 to 512 readings, but I find 16 is usually enough to filter out any jitter or noise for my use.

[RavensFan]

---

@chrissheffuni wrote:

How quickly do you need to acquire the data? if its not a very dynamic test have you considered oversampling and averaging your data? The NI9205 can acquire at 250ks/S

---

Do you realize you are replying to a message that is nearly 5 years old?

[chrissheffuni]

Actually I didn't. But do you realise you're also replying to a question that's 5 years old???

[RTSLVU]

---

@RavensFan wrote:

---

@chrissheffuni wrote:

How quickly do you need to acquire the data? if its not a very dynamic test have you considered oversampling and averaging your data? The NI9205 can acquire at 250ks/S

---

Do you realize you are replying to a message that is nearly 5 years old?

---

Yet another reason why ALL threads should be locked after a determined amount of time

[chrissheffuni]

Such a shame that there are such know it all's in the developer zone. One solution isn't always the best solution and even though a solution has been offered, over time new solutions and methods come to light.  This thread maybe old, however the question is always asked from every newcomer to lab view.  How do I smooth the data on a voltage signal? Because the threads are left open the newcomer can explore every answer and choose the solution that best suits their needs. Even though this is my first answer on devzone. I have been developing lab view apps for 10 years plus I spend 50 to 75 k every year on NI hardware and program in Labview for windows Real time and FPGA. They are for test equipment for leading automotive and aerospace companies.  So get off your high horse mate and realise that there are many new people coming to lab view who may find my answer useful even after 5 years!  

[JÞB]

---

@chrissheffuni wrote:

Such a shame that there are such know it all's in the developer zone. One solution isn't always the best solution and even though a solution has been offered, over time new solutions and methods come to light.  This thread maybe old, however the question is always asked from every newcomer to lab view.  How do I smooth the data on a voltage signal? Because the threads are left open the newcomer can explore every answer and choose the solution that best suits their needs. Even though this is my first answer on devzone. I have been developing lab view apps for 10 years plus I spend 50 to 75 k every year on NI hardware and program in Labview for windows Real time and FPGA. They are for test equipment for leading automotive and aerospace companies.  So get off your high horse mate and realise that there are many new people coming to lab view who may find my answer useful even after 5 years!  

---

That is uncalled for.

A new thread would point out the outdated advice in these related threads and incorporate the newer pt by pt vis as an un-masked solution that could be marked by a new commercial.  You can't mark threads you don't start (unless you are gifted)

Its easy to get sidetracked to an old thread while researching a problem.   Pointing it out to the poster is kind and helps others that may read a thread that has been accidentally bumped beyond its ops lifetime. 

[JÞB]

---

@RTSLVU wrote:

Also consider a "running average"...

Say you are taking a measurement every 100mS

Average the reading over 1 second (10 readings) and save or display the average.

Most power analyzers do this and offer several levels of averaging depths. One of my Yokogawa power analyzers will average from 2 to 512 readings, but I find 16 is usually enough to filter out any jitter or noise for my use.

---

Use pt by pt  mean.vi

[chrissheffuni]

The older advice is not necessarily outdated. My comment is totally caled for. I looked at the thread to see if I could find a better way of smoothing a voltage signal. I left leaving a little bit of advice which might help someone in the future. I then get a knowitall telling me how old the thread is and then copying my comments into another thread.  

Wont be coming on her again is just like any other social media. ... 

[RTSLVU]

---

@chrissheffuni wrote:

Such a shame that there are such know it all's in the developer zone. One solution isn't always the best solution and even though a solution has been offered, over time new solutions and methods come to light.  This thread maybe old, however the question is always asked from every newcomer to lab view.  How do I smooth the data on a voltage signal? Because the threads are left open the newcomer can explore every answer and choose the solution that best suits their needs. Even though this is my first answer on devzone. I have been developing lab view apps for 10 years plus I spend 50 to 75 k every year on NI hardware and program in Labview for windows Real time and FPGA. They are for test equipment for leading automotive and aerospace companies.  So get off your high horse mate and realise that there are many new people coming to lab view who may find my answer useful even after 5 years!  

---

The old threads will still be there and when a newb asks the same old question but slightly different they can reference the closed thread and say "I tried everything in this thread".

[chrissheffuni]

So let me get this right. You are saying that it is easier to trawl through many different threads all with similar titles but not the same using a search engine. Some of which are locked. To eventually find the solution, or not (depending on the text used in the search)? Rather than look at THE question and then scroll through all suggested answers and solutions to then pick the right one for that application and maybe add some useful information at the same time?

I see that's where LabVIEW help is all wrong, it has all the content relating to the same device or method of coding all in the same place! Can I ask, when you code in LabVIEW do you have many different pieces of code all doing the same thing but for different parts of the application???? Could I suggest you sit the LabVIEW core 1-3 you may change your mind!