Chrono Calibration for the Tunnel
Posted 10-21-2009 at 05:01 PM by HuskerP7M8
Updated 10-23-2009 at 04:28 PM by HuskerP7M8 (Stupid errors in my math!)
Updated 10-23-2009 at 04:28 PM by HuskerP7M8 (Stupid errors in my math!)
Since the absolutely horrid weather I’m experiencing has curtailed harvest and I’ve become incredibly bored, I decided to install the infrared emitters on my two CED M2 chronographs and conduct some evaluation and calibration testing.
Because I was unable to shoot outdoors with a rimfire, I decided to use a semi-automatic BB gun and did the testing in my garage with the lights off. This testing will be repeated in the tunnel at rimfire velocities but I wanted to try and satisfy my curiosity about obtaining reliable data with consumer grade chronographs.
I placed the chronos upside down using the emitters as a flat stable base on a 5’ length of plywood with #1 touching #2 and then I numbered the chronos with a magic marker as #1 and #2 based on their orientation from the shooter’s POV.
The CED M2 has a 24” screen spacing and when I measured the distance between the sensor slots....I found that chrono #1 was a full 3/16” longer than chrono #2.
Based on the ratio of 3/16” to 24” (0.0078125), that meant chrono #2 should read 2.15 ft/sec faster at my airgun average velocity of 275 ft/sec.
Amazingly, after firing 25 shots thru a #1-#2 configuration of the chronographs and 25 shots thru a #2-#1 configuration while including my estimate of the BB’s loss of velocity over 4’.....real test data almost exactly matched my prediction and now allows me to apply a calibration factor so that the velocity of both chronos agrees.
BTW I’m not concerned with how closely the two chronos match true/actual velocity (I’m going to assume they’re close based on the manufacturer’s claims). I’m primarily concerned with calibrating them to read the same so as to gather reliable data for velocity loss at various distances.
This calibration factor which results in a 2.15 ft/sec velocity reduction for chrono #1 may seem very small but if you extrapolate to a rimfire velocity of say 1050 ft/sec....it increases to 8.19 ft/sec.
Next, I needed to determine the variance in readings between the two chronographs due to the variables I can’t measure. My testing data showed a standard deviation of 0.376 ft/sec, which means an uncertainty of 0.752 ft/sec either side of the mean for 95% confidence of velocities in the 275 ft/sec range.
Again, this seems small but it increases to 2.870 ft/sec at a velocity of 1050 ft/sec.
Some Conclusions:
1. The 2.870 ft/sec error is proportionately large relative to the velocity loss over 50 yards (50-60 ft/sec), so calculating a Ballistic Coefficient will require it be stated as a range of values and not as a number accurate to three decimal points.
Example: A calculated BC of .140 could range from .136 to .144.
It’s also important to remember that BC’s vary with the velocity a projectile is tested at “and” the extreme changes in drag within the transonic velocity range result in large changes for BC values. All in all though....this ain’t bad and I might be able to get some interesting numbers for comparison.
I do have the option of throwing away the 24” brackets and using longer screen spacings in an attempt to improve accuracy. Maybe I’ll try that, but I’ll have to remember that since the CED M2 has no provisions for screen spacing options, I’ll have to calculate a calibration ratio for true velocity.
2. The 2.87 ft/sec uncertainty should only result in 0.03” movement at 50 yards based on the “ballistics only” and that will allow some very accurate calculations for the majority of testing I have in mind. I verified this with both Professor Art Pejsa's ballistic software and Bryan Litz's "Point Mass Ballistics Solver".
None of the above has been rechecked for accuracy/mistakes or incorrect assumptions regarding the correct procedures needed for this type of testing, but since I plan on redoing it all at a later date with rimfire velocities.....I’m done for now.
Landy
PS I’ve now re-edited this twice and I hope it’s finally correct but I seem to be making a multitude of errors lately and can’t/won’t guarantee I’ve finally got it right. LOL
Because I was unable to shoot outdoors with a rimfire, I decided to use a semi-automatic BB gun and did the testing in my garage with the lights off. This testing will be repeated in the tunnel at rimfire velocities but I wanted to try and satisfy my curiosity about obtaining reliable data with consumer grade chronographs.
I placed the chronos upside down using the emitters as a flat stable base on a 5’ length of plywood with #1 touching #2 and then I numbered the chronos with a magic marker as #1 and #2 based on their orientation from the shooter’s POV.
The CED M2 has a 24” screen spacing and when I measured the distance between the sensor slots....I found that chrono #1 was a full 3/16” longer than chrono #2.
Based on the ratio of 3/16” to 24” (0.0078125), that meant chrono #2 should read 2.15 ft/sec faster at my airgun average velocity of 275 ft/sec.
Amazingly, after firing 25 shots thru a #1-#2 configuration of the chronographs and 25 shots thru a #2-#1 configuration while including my estimate of the BB’s loss of velocity over 4’.....real test data almost exactly matched my prediction and now allows me to apply a calibration factor so that the velocity of both chronos agrees.
BTW I’m not concerned with how closely the two chronos match true/actual velocity (I’m going to assume they’re close based on the manufacturer’s claims). I’m primarily concerned with calibrating them to read the same so as to gather reliable data for velocity loss at various distances.
This calibration factor which results in a 2.15 ft/sec velocity reduction for chrono #1 may seem very small but if you extrapolate to a rimfire velocity of say 1050 ft/sec....it increases to 8.19 ft/sec.
Next, I needed to determine the variance in readings between the two chronographs due to the variables I can’t measure. My testing data showed a standard deviation of 0.376 ft/sec, which means an uncertainty of 0.752 ft/sec either side of the mean for 95% confidence of velocities in the 275 ft/sec range.
Again, this seems small but it increases to 2.870 ft/sec at a velocity of 1050 ft/sec.
Some Conclusions:
1. The 2.870 ft/sec error is proportionately large relative to the velocity loss over 50 yards (50-60 ft/sec), so calculating a Ballistic Coefficient will require it be stated as a range of values and not as a number accurate to three decimal points.
Example: A calculated BC of .140 could range from .136 to .144.
It’s also important to remember that BC’s vary with the velocity a projectile is tested at “and” the extreme changes in drag within the transonic velocity range result in large changes for BC values. All in all though....this ain’t bad and I might be able to get some interesting numbers for comparison.
I do have the option of throwing away the 24” brackets and using longer screen spacings in an attempt to improve accuracy. Maybe I’ll try that, but I’ll have to remember that since the CED M2 has no provisions for screen spacing options, I’ll have to calculate a calibration ratio for true velocity.
2. The 2.87 ft/sec uncertainty should only result in 0.03” movement at 50 yards based on the “ballistics only” and that will allow some very accurate calculations for the majority of testing I have in mind. I verified this with both Professor Art Pejsa's ballistic software and Bryan Litz's "Point Mass Ballistics Solver".
None of the above has been rechecked for accuracy/mistakes or incorrect assumptions regarding the correct procedures needed for this type of testing, but since I plan on redoing it all at a later date with rimfire velocities.....I’m done for now.
Landy
PS I’ve now re-edited this twice and I hope it’s finally correct but I seem to be making a multitude of errors lately and can’t/won’t guarantee I’ve finally got it right. LOL
Total Comments 0
