next topic »

[on:]

My project is a shield for the Arduino.  It is designed to monitor engine knock voltage from peizo sensor w/ conditioning circuit, determine RPM via a frequency to voltage convertor, determine when knock is present by verification against RPM, datalog on serial, and display on a LCD.

I am currently considering a s/w problem because I wrote the code.  I think I did a pretty good job on my code, it's structure, and comment it, but there must be a problem with it somewhere.  Any help is greatly appreciated!  Edit: The voltages do not even remotely reflect the actual voltage as measured by a meter.

Here is the code: http://www.squirrelpf.com/knockboard/code.txt

Thanks,
Frank

[Reply #1 on:]

Well you might start by telling us what works and what doesn't work  ;D

Lefty

[Reply #2 on:]

Wow... nicely done Frank. LOL.  Fixed description.

[Reply #3 on:]

So, what was the problem?  How did you fix it?  It's useful to know, because others will search these forum postings for help when they have the same kind of trouble.

[Reply #4 on:]

The problem isn't fixed.  i just fixed my post which didn't describe what was wrong.

[Reply #5 on:]

Hm.  It "does not remotely reflect the voltages" the way you expect.

How about telling us more?  What's producing the voltage?  What range of voltages do you read with a meter?  Where are you sending that signal?  How are you reading that signal?  What range of readings do you get?  Is the reading high when the meter shows a high voltage?  Is the reading low when the meter shows a low voltage?  Are the readings noisy?  random?  clipped?  inverted?  not the same number you see on the meter?  ...

How to ask questions the smart way.

[Reply #6 on:]

Well I figured people would give it a try.  Put to run it down, 5 volts would appear around 1 volt and it tens to fluctuates a bit.  If I change it to 4 volts, it would still be just a bit less then 1 volt.

[Reply #7 on:]

Any ideas?

[Reply #8 on:]

I suspect you have a issue of mixing byte with floating point.

byte    knock_volt;                         // this is the raw knock voltage at any instance in time

0-255 is all that knock_volt will ever contain. It will rollover to 0 when it reaches 256.


 knock_volt = analogRead(knockPin);                               // grab the sensor data for the knock sensor
  knock_volt = map(knock_volt, 0, 1023, 0, 255);                   // remap the 10-bit input to 8-bit


Okay, you even map the 10 bits down to 8.
May be better to keep the 10 bits and use INT as you lose precision.


 Serial.print(knock_volt, DEC);
    Serial.print("\t");
    

That should show a value between 0-255?



  printDouble(knock_volt * 0.019531, 3);                                                  // Knock Volts

- knock_volt is not a double
- knock_volt * 0.019531 ?

  255 the maximum knock_volt can be * 0.019531 is never going to be more than 4.

I suggest you do the scaling completely with integer (x10 or x100), then insert the decimal for printing.

For example...
knock_bits  0-1023 or 0-5V

knock_bits = 511 or 2.5V or  204.4 bits per volt.
since we want two decimal place precision voltsPerBit = 2044

long volts (because int will overflow)

volts = (knock_bits * 1000) / 2044


volts now contains the actual volts * 100
Very fast, no floating point.
It's just a matter of formatting the output.

Avoid using floating point.
When reading only 1024 bits, the precision is ony 1 bit or 1/1000 easy to handle without floating point.


Hope that helps.

[Reply #9 on:]

As George says, you're losing data in knock_volt.

Code:

byte knock_volt;
knock_volt = analogRead(knockPin);
knock_volt = map(knock_volt, 0, 1023, 0, 255);

After this, the highest value you could have in knock_volt is 63.  What's worse, the value 63 will appear when the voltage is at 1/4, 1/2, 3/4 and full levels.

[Reply #10 on:]

Thanks folks!  That took care of it!    

Lesson learned!