Sampling rate for sensor

[umashankar]

Hello everyone,

i got the raw data from potentionmeter and wheelspeed but there are a lot of noise in it making analyzing data impossible
that's why i want to know what is the ideal frequency to apply filter or log data on for different sensor
thank you in advance

[jd74914]

I have a few questions for you relevant to your questions:

1) How fast are you sampling now?
2) What filters are you using and how are you applying them?
3) Are your sensors noisy because of grounding, shielding, or connection issues (ie: are they noisy when the car is static)?
4) Can you post a sample picture of your data?

[Claude Rouelle]

If I tell you 10 HZ for your damper potentiometers, would you believe me? If I tell you 500 HZ, would you believe me?

If one day a design judge ask you that question and you answer "X Hz...because another design judge told me so", do you think you would be credible?

This is the kind of question that pisses me off. Or at least that is the way the question is formulated that is upsetting. Have you done your own research?

Here is a bit of help that will help you to THINK before you ask
- Do you think it is necessary to log your engine water and oil temperature at 500 Hz?
- In fact isn't one of the questions "how often does your water temperature changes"?
- Similarly how often is your damper changing its position?
- What about natural, undamped suspended mass frequency as an indication?
- Or some basic, realistic succession of road bumps simulation with a basic wheel rate, tire rate, linear and symmetrical damping (same at low speed and high speed, same in bump and rebound)
- Never heard about the Nyquist frequency?
- Where could the noise come from? Electromagnetic interference (I have see some strange sensors noise if the sensor is close to the ignition coil and the senors and wires are not well insulated)? Issue with no grounding?
- I have learnt from 7 post rig testing that the bracket that holds a sensor should have a natural frequency of at least 250 Hz. I see in Formula Student brackets made of bend sheet of steel that must vibrate more than the sensors itself!
- What is your sensor resolution? Do you try to measure 40 mm of damper movement with a sensor made to measure 100 mm of stroke (in other words use 2 of the 5 volts)?
- Do you realize what a high or unnecessary high logging frequency will do on the first (speed) and second (acceleration) derivative of your damper measured movement?
- How did you calibrate your sensors: 2 points, 5 points, 10 points, blindly follow the sensor spec sheet?
- What about looking at a frequency distribution or Power Spectrum density to see where most of the frequencies are?
- Here is an idea: put 2 sensors one above the other parallel to your damper. Log one at 50 Hz and the other at 250 Hz. Compare and analyze: what are your conclusions?


For God sake instead of asking "what is the ideal logging frequency " could you simply ask "what are the criteria to chose the ideal logging frequency" and come with a beginning of answer on your own.
Otherwise you will be cataloged as another "Sir give me the answer". This is Formula Student, not a exam similar to the ones I see in many Indian Universities.

[Jay Lawrence]

umashankar,

Please don't be put off by Claude. It would suck to see people give up because they were put down by people on this forum who supposedly knew everything the day they were born...

Come back with a bit more information and you'll probably find that you'll get a better response

[Z]

I agree with Claude on this one.

Z

[Claude Rouelle]

Jay,

Have you ever spent a few weeks in India?

Claude

[Jay Lawrence]

No Claude I haven't, but the country of origin of the OP had nothing to do with my comment. I understand that there are issues with authority figures and current educational systems in India, and this may have a negative affect on students there. However, you (and Z) have not spent your whole lives knowing everything that you currently know, but your posts are often condescending enough to suggest otherwise. Jim asked some good questions and then you asked some more good questions, but mixed it up with superiority complex bullshit.

At the end of the day, your power trip is less important than the growth of people like the OP.

[Bemo]

Having been at an Indian competition myself, I know pretty well, what Claude is talking about. Most Indian students are only looking for someone telling them what to do and at some point you just get sick of that. Therefore you don't tell them the answers to their questions anymore, instead you try to make them think about the issue themselves first.

This has nothing to do with stupid prejudices, it is just a fact due to a poor education system and so on. But as they aren't taught how to properly approach engineering problems, they need guidance from somewhere else. In my opinion, the response Claude gave is much more worth in the long term than just giving the answer to the actual question (if the right conclusions are made).

[Claude Rouelle]

Jay,

Condescending, Superiority? No. I have made many, many mistakes and I still do but I learn and I a still learning from them. Provocative? Yes.

You have no idea of the amount of personal time (free of charge) I give through this forum, though emails and in the FS paddocks to students. But after a while when you realize the system is not working and the advice given to the students have practically no effect you need to attack the system.
With some people, and against my nature, kindness has no effect. So I feel I need to provoke. My experience has shown it helps. You don't make omelets without braking eggs.

[mech5496]

Z has hacked Claude's account :P

Jokes aside, I agree with you. It is hilarious to see the requests coming through the teams' e-mail or facebook page, not asking guidance or thoughts, but the exact answer.