Data logging sample rate ???

[WildBro]

I am setting up a data logger on a turbo bike and was wondering if setting the sample rate to high will lead to problems?  I have a lot of room to store data, so that would not be a problem.  I was just thinking if the frequency is to high I may be over whelmed interperting it.

A run at Maxton will be about 20sec, over the last 1/4 mile I guessed a sample rate of 30Hz on a channel will sample each foot traveled.  Is that to much sampling?  Logger can do a total of 2000Hz, I am at 331HZ now with channels still open (not used). (the 331Hz gives me 7 hours of logging)

Thanks for any help and ideas,
Bill

[Rex Schimmer]

Bill,
What things are you going to look at? You don't want to try to log data faster than your sensors can work but 30 HZ is really pretty slow. I do alot of data logging on hydraulic systems and typically try to look at collecting data every milisecond, that would be 1000 HZ. If you have good transducers I would suggest running as fast as you can, if you turn your engine 10 grand then it will have rotated 5.5 times between your data points, hell you could have a rod out the side in that time and never knew what happened! At 1000 HZ you are looking at your data 6 times ever revolution.

Rex

[Harold Bettes]

Rex,

What kind of data logger do you use on the hydraulic systems?

I am looking for a simple data logger that is not too expensive so that I can look at some trend data. Not particularly interested in absolutes.

Regards,
HB2

[Sumner]

Rex,

What kind of data logger do you use on the hydraulic systems?

I am looking for a simple data logger that is not too expensive so that I can look at some trend data. Not particularly interested in absolutes.

Regards,
HB2

Since he didn't have to pay for it I'll bet it was a lot.

Look at the megaSquirt. Besides controlling the EFI you can data log a lot of inputs with it even if you aren't using it with EFI.  The cost can be under $300 plus whatever sensors you will use.

c ya,

Sum

[Rex Schimmer]

Harold,
I have a 16 channel Dataq system with mega buck transducers and it takes a big suit case to carry it around. I don't use it! I bought a USB Insturments DS1M12 "Stingray" two channel data logger. $200 ! It likes 0-5 volt dc signal and you can program it to sample up to a million times per second. Pretty neat piece. It does need to be hooked to a lap top to work. I think their web site is www.USB-Instruments.com. You will find that the transducers cost more than the data logger!

Rex

[smcleod007]

I am setting up a data logger on a turbo bike and was wondering if setting the sample rate to high will lead to problems?  I have a lot of room to store data, so that would not be a problem.  I was just thinking if the frequency is to high I may be over whelmed interperting it.

A run at Maxton will be about 20sec, over the last 1/4 mile I guessed a sample rate of 30Hz on a channel will sample each foot traveled.  Is that to much sampling?  Logger can do a total of 2000Hz, I am at 331HZ now with channels still open (not used). (the 331Hz gives me 7 hours of logging)

Thanks for any help and ideas,
Bill

Hi Bill, 

These are the rates I'm starting with, on an Aimsport MXL PRO Logger. I did a quick screen print of Race Studio 2 from my desktop computer instead of the laptop computer which plugs into the bike so the Lambda and GPS didn't show up. Some of the sensors you can select from the drop down lists on the Aimsports loggers will automatically be limited to they're highest recording rate. This is so you can't assign a 500Hz sampling rate to a sensor that cant physically exceed 100Hz and so on. I plan on setting my recording rates on the high side for the first couple of passes. Once I see a bunch of redundant data on a channel that doesn't chance very quickly I'll reduce the sampling rate. I would like to install a lot more sensors to help analyze what the chassis does at speeds over 200, but first I'm trying to save up for a bigger turbo system with an intercooler. The GT25 I'm currently using is pretty much maxed out.

[WildBro]

Hi Scott,
That's a good idea.  I'll max out the sample rate for each sensor and go back and turn it down after seeing common data.
Talk to you soon.

Bill
P.S.  Almost done with the bike(1st stage), just a few more parts to make

[ddahlgren]

You have to sample at least twice the frequency of the event you are trying to capture. Examples might be for minimums.. shocks 200 hz  tps map rpm oil pressure etc 50 hz Engine temp oil temp air temp things that move slowly 5Hz.  It is called the Nyquist frequency do a web search and some reading and will all make sense.
Dave