Minimum sampling rate for DIY datalogger?
 

I'm planning to build an arduino datalogger as a winter project.
I want to log the following.

• Boost/Vacuum.
• Wideband A/F ratio
• RPM
• G forces
• Knocking
• Option of logging speed. For use on track, and disabled on the road. No need to record potential evidence ;)

I have already ordered all the sensors needed. I believe knocks and rpm signals can be taken directly from the DME. Can I get the speedometer from the DME as well?

I believe the challenge will be logging knocks and A/F ratio as these require the fastest sampling rate to avoid missing an event. What would be the slowest sampling rate I could run and still be fairly sure of not missing anything?

I would like to do the same thing and have been looking for products to monitor and log the same parameters that you are. Would you be willing to provide information about the set up that you are attempting to make? How much do you expect to spend on it in total?

I'm expecting to spend 80- 150 dollars depending on which route I go.
I don't mind sharing info. Arduino is open source anyways, so why shouldn't I share.

Right now the question is if it is even possible to do this on the arduino platform. I won't know until I've determined the minimum sampling rate needed.

I would say your best bet is to contact Rogue Ant. He is a pro when it comes to 951 and DIY stuff like this.

Tuning in...

Which Arduino board are you using?

WBO2 are fairly slow, 10Sa/s is probably sufficient. 50Sa/s would definitely be sufficient for everything besides the knocking signal...
The knock signal is very fast. But you could configure the Arduino to use this input to trigger an external interrupt... And log knock counts using the interrupt.


-Rogue

Hi Rogue.
I'm completely new to DIY electronics, so I appreciate all and any suggestions you might have. I have bought the Nano board, but my mate and electronics mentor has got the Mega. That way we can try some different approaches.

I didn't think of using the knock signal as a trigger. Makes it a lot simpler.

What info is available directly from the DME? Anything else that could be fun to datalog?

The Nano board should be good for this :)
There is only 8 ADC in the DME, so you could log pretty much all of them (you would have to tap the lines externally to the DME). The ADC lines are (only 7 are used):

AFM
Volt
AIT
Coolant
Baro
TPS
FQS

Obviously you won't be able to log the FQS without tapping the digital board in the DME. But the other 6 you should be able to log. The internal variables to the DME could only be retrieved using the DME serial output... However that requires either re-activating the serial code, or writing new serial code on the DME EPROM.

-Rogue

Great.
Forgive my ignorance buy what is AIT amd FQS?
Also. How is the speedometer driven on the late cars?

AIT = Air Intake Temp
FQS = Fuel Quality Switch (the 8-position rotary switch on the DME)

Speedo, IIRC, is a VR signal from the transmission, fed to the DME which then outputs a correct square wave signal to the gauge...

See, I knew Rouge would be there with answers. I can only hope to understand half that you do about the electronics on our cars.

I try to help, when I can :)

Bump, any progress?

I've been swamped at work, but I've managed to get the LCD and the pressure sensor going. Whenever I have time I work on the menu system, but I need a good rotary encoder with push button to test it out. I've been looking for one on eBay, but I need one with threads around the collar for secure fastening, and they seem hard to find.

The 3 axis accelerometer is sitting on my desk only lacking some high quality ribbon cable suitable for an automotive installation. The hard drive cable I tried breaks too easily once soldered.

Any tips on where I can get high quality ribbon cable of a lower gauge than HDD cable?

What I really need is time, but I will get plenty of that soon. My car needs some work before I can start testing it in the car anyway, so I'm not in a hurry.

Hey Rogue.
Do you know if there are lots of interference problems when tapping the DME? I'm just trying to think ahead and plan any signal filtering I might need.

# oost/Vacuum.
# Wideband A/F ratio
# RPM
# G forces
# Knocking
in that types of technique many interference problem am agree with bebbetufs signal filtering needing here

What HDD cable are you using? The UDMA cables are 28gauge, and it is going to be hard to find ribbon cables with smaller gauge than that.

Here is a link to some rotary encoders, and ribbon cables:

Rotary Encoder

Ribbon Cables

Signal filtering will depend on your sampling frequency...

If you have a decent sample rate, then you might want to do a simply averaging filter so your display values aren't jumping around.

Knocking event signal is very fast, I would not apply a filter to this signal - you want to catch each event.

Thanks. Rotary encoder is sitting in my breadboard and I have found some 24g ribbon cable at Farnell. I might buy some of that, For now I'm only using HDD cable for bench testing. I'm also considering shielded VGA cable, but I'm unsure of their suitability.

I'm mostly concerned with electrical interference which could affect the electronics and cause loss of data or other failures. I'm wondering if I will need electrical filtering of the signal lines from the DME to keep interference out of the Arduino.

I will apply a 5 sample average to most of my values anyway, and I will use an external interrupt to capture the knocking signal.

The what is the ADC impedance on your Arduino board? Most likely it is quite high - and if so, then there shouldn't need any filtering.