Author Topic: Calculating resolution  (Read 2515 times)

0 Members and 1 Guest are viewing this topic.

Offline ElectricGuyTopic starter

  • Regular Contributor
  • *
  • Posts: 240
  • Country: pt
Calculating resolution
« on: March 06, 2015, 04:06:03 am »
Hey everyone;

Help one on this one, because i'm not getting there alone.

I have a system that measures the velocity across two points. Thoses points are separated from each other 30cm.
My PIC micro starts a timer when the projectil travels across the first sensor and stops on the second sensor.
My Timer is configured to tick every 4uS (8Mhz/4/(1:8 prescaller)-1.

In software i'm multiplying my timer value by 4 so i can get the time traveled in uS.
So, my velocity in m/s is V=300000/(timervalue*4)

All is working fine great, i get the velocity. but now i want to figure the resolution of the system, and i'm not getting there.

Can somebody help out?
Thank you!
Thank you!
Regards
ElectricGuy
 

Offline KerryW

  • Regular Contributor
  • *
  • Posts: 112
  • Country: us
Re: Calculating resolution
« Reply #1 on: March 06, 2015, 04:40:54 am »
Your resolution is 3uS.  You probably want to know the resolution in M/S, though. don't you?

The resolution in M/S changes with the velocity.

1 tick = 75000 M/S, 2 ticks = 37500 M/S, resolution = 37500 M/S

10000 ticks = 7.5 M/S, 10001 ticks = 7.49925 M/S, resoluyion = 0.00075 M/S
One accurate measurement is worth a thousand expert opinions
- Adm. Grace Hopper
 

Online IanB

  • Super Contributor
  • ***
  • Posts: 11891
  • Country: us
Re: Calculating resolution
« Reply #2 on: March 06, 2015, 04:46:16 am »
If you are measuring time intervals, your resolution is 4 µs (the smallest resolvable difference between two measurements is 4 µs).

If you are measuring speed, the resolution is a function of the time resolution dt, the distance between measurement points s and the speed you are measuring, v. Your speed resolution is given by:

    dv = [s / t] - [s / (t + dt)]

where

    t = s / v

hence

    dv = v * [1 - s / (s + v * dt)]

Therefore the resolution is better at lower speeds, greater distances, and smaller clock intervals.
 

Offline ElectricGuyTopic starter

  • Regular Contributor
  • *
  • Posts: 240
  • Country: pt
Re: Calculating resolution
« Reply #3 on: March 06, 2015, 04:53:23 am »
Ok, i understand now. And yes, i want the m/s resolution.

Thank you IanB and KerryW.

Actually i was getting those values, but didn't understand them.

I think for each measure i will calcule and print also the resolution. So i can always know the resolution that i'm getting.
« Last Edit: March 06, 2015, 04:55:34 am by ElectricGuy »
Thank you!
Regards
ElectricGuy
 

Offline babysitter

  • Frequent Contributor
  • **
  • Posts: 893
  • Country: de
  • pushing silicon at work
Re: Calculating resolution
« Reply #4 on: March 06, 2015, 05:34:40 am »
Hi,

your resolution with this timing setup will always be the same as long as you don't change one of the contributing values sensor distance or timer settings. I think it would be more interesting to printout the interval you dont know about in m/s, like in Speed between X1 m/s and X2 m/s. X1 is the corresponding speed according to the timer value, X2 the speed that corresponds to timer-1, as the actual trigger event is inbetween those two, not earlier than timer-1 but before timer.
I'm not a feature, I'm a bug! ARC DG3HDA
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf