LTI 20-20 UltraLyte 100 Calibration checks

LTI 20-20 UltraLyte 100 Calibration checks

Author
Discussion

Mill Wheel

6,149 posts

195 months

Wednesday 25th November 2015
quotequote all
tapereel said:
The speedometer doesn't work like that.
The speed gun really is a speed gun, it measures and indicates distance only after the speed has been measured.
Perhaps it is more accurate to say "the speed gun isn't a distance gun, it's a speed gun that can also measure distance."
I think what you meant to say was that it sends out variable power laser pulses (ensuring the receiver isn’t saturated or under-driven, by means of pulse width measurement).
The reflections are received by the avalanche diode (adjustable bias) which can only measure intensity with time, not Doppler shift or any other effect; hence it cannot distinguish reflections from static or swept targets.

This signal is sent to a comparator (reference level adjustable for background levels and countermeasure noise), which removes the intensity information to give ‘clean’ digital pulses, which are in turn is sent to the gating (for predictive ‘lock on’), interpolation and timing circuits.
The rising edge delay for each returned pulse is measured against an internal reference pulse (processed in much the same way). Therefore, measurement of a moving target and sweeping a stationary continuous surface (not necessarily flat, depending on hand movement) will result with the same digital pulse train from the comparator, each consecutive pulse having a uniform rate of change of delay with time.

Simply put, the device is designed to be used as a surveying tool, where it very accurately measures distances (and is used as such by some police accident investigators to produce an accurate record of a crash scene).
Additional Software is employed to use those very accurate range checks, to determine the speed of a vehicle. It does this by interpreting the changing DISTANCE from the device to the vehicle over a set period of time, to calculate the speed.

The range checks of course involve MOVING the speed meter to track the target, and it is this which can introduce errors in the distance measurements - alluded to by Rovinghawk.

It really isn't a speed gun - it was adapted to become a speed gun by the addition of software.


V8LM

5,166 posts

208 months

Wednesday 25th November 2015
quotequote all
Mill Wheel said:
I think what you meant to say was...
Unfortunately he refuses to say anything other than that everyone else is wrong, or, I believe, complete bks such as "The speed reading however will be exactly right no matter what the clock rate is".


Edited by V8LM on Wednesday 25th November 18:34

Pete317

1,430 posts

221 months

Wednesday 25th November 2015
quotequote all
V8LM said:
Mill Wheel said:
I think what you meant to say was...
Unfortunately he refuses to say anything other than that everyone else is wrong, or, I believe, complete bks such as "The speed reading however will be exactly right no matter what the clock rate is".


Edited by V8LM on Wednesday 25th November 18:34
Actually, he's right about the speed being right, except that the other things he says leads me to believe that he doesn't know why he's right, and is probably just parroting what someone has told him.

If, for example, the timing clock is running slow, then it will measure the pulse propagation time as less than it actually is.
But, because the rate at which the pulses are sent is controlled from the same clock, the pulses will also be sent at a slower rate.
This means that the inaccuracy in the measured pulse propagation time will be canceled out by the fact that the time between the pulses will also be longer, and so the difference between the two pulse delays will work out the same, regardless of the inaccuracy of the clock.

So yes, an inaccuracy of the clock does not affect the speed reading, but it doesn't alter the fact that the device works by measuring distance and converting the relative distance into speed. The fact that it may not measure the distance accurately is immaterial, it still measures the distance but it does not need to measure the absolute distance accurately in order to get an accurate speed reading.

But, as it turns out, crystal oscillators are so accurate that any errors would be slight anyway, even if not compensated for. Unless of course something went wrong with the electronics, in which case the clock may be wildly wrong - but such devices are normally equipped with watchdog circuits which detect when the clock is out of spec.

Edited by Pete317 on Wednesday 25th November 20:56

V8LM

5,166 posts

208 months

Wednesday 25th November 2015
quotequote all
Pete317 said:
Actually, he's right in the bit about the speed being right, except that the other things he says leads me to believe that he doesn't know why he's right, and is probably just parroting what someone has told him.

If, for example, the timing clock is running slow, then it will measure the pulse propagation time as less than it actually is.
But, because the rate at which the pulses are sent is controlled from the same clock, the pulses will also be sent at a slower rate.
This means that the inaccuracy in the measured pulse propagation time will be canceled out by the fact that the time between the pulses will also be longer, and so the difference between the two pulse delays will work out the same, regardless of the inaccuracy of the clock.

So yes, an inaccuracy of the clock does not affect the speed reading, but it doesn't alter the fact that the device works by measuring distance and converting the relative distance into speed. The fact that it may not measure the distance accurately is immaterial, it still measures the distance but it does not need to measure the absolute distance accurately in order to get an accurate speed reading.

But, as it turns out, crystal oscillators are so accurate that any errors would be slight anyway, even if not compensated for. Unless of course something went wrong with the electronics, in which case the clock may be wildly wrong - but such devices are normally equipped with watchdog circuits which detect when the clock is out of spec.



Edited by Pete317 on Wednesday 25th November 20:43
Not quite. Such independence in clock would require the speed of light to be dependent on the clock too.


Edited by V8LM on Wednesday 25th November 21:42

Pete317

1,430 posts

221 months

Wednesday 25th November 2015
quotequote all
V8LM said:
Pete317 said:
Actually, he's right in the bit about the speed being right, except that the other things he says leads me to believe that he doesn't know why he's right, and is probably just parroting what someone has told him.

If, for example, the timing clock is running slow, then it will measure the pulse propagation time as less than it actually is.
But, because the rate at which the pulses are sent is controlled from the same clock, the pulses will also be sent at a slower rate.
This means that the inaccuracy in the measured pulse propagation time will be canceled out by the fact that the time between the pulses will also be longer, and so the difference between the two pulse delays will work out the same, regardless of the inaccuracy of the clock.

So yes, an inaccuracy of the clock does not affect the speed reading, but it doesn't alter the fact that the device works by measuring distance and converting the relative distance into speed. The fact that it may not measure the distance accurately is immaterial, it still measures the distance but it does not need to measure the absolute distance accurately in order to get an accurate speed reading.

But, as it turns out, crystal oscillators are so accurate that any errors would be slight anyway, even if not compensated for. Unless of course something went wrong with the electronics, in which case the clock may be wildly wrong - but such devices are normally equipped with watchdog circuits which detect when the clock is out of spec.



Edited by Pete317 on Wednesday 25th November 20:43
Not quite. Such independence in clock would require the speed of light to be dependent on the clock too.


Edited by V8LM on Wednesday 25th November 21:42
You're missing something - or maybe I'm not explaining it as clearly as I should.
A shortness in the perceived delay is exactly compensated for by the longer actual time between consecutive pulses, which makes the perceived difference between the two delays correct.


Edited by Pete317 on Wednesday 25th November 21:57

V8LM

5,166 posts

208 months

Wednesday 25th November 2015
quotequote all
Pete317 said:
V8LM said:
Pete317 said:
Actually, he's right in the bit about the speed being right, except that the other things he says leads me to believe that he doesn't know why he's right, and is probably just parroting what someone has told him.

If, for example, the timing clock is running slow, then it will measure the pulse propagation time as less than it actually is.
But, because the rate at which the pulses are sent is controlled from the same clock, the pulses will also be sent at a slower rate.
This means that the inaccuracy in the measured pulse propagation time will be canceled out by the fact that the time between the pulses will also be longer, and so the difference between the two pulse delays will work out the same, regardless of the inaccuracy of the clock.

So yes, an inaccuracy of the clock does not affect the speed reading, but it doesn't alter the fact that the device works by measuring distance and converting the relative distance into speed. The fact that it may not measure the distance accurately is immaterial, it still measures the distance but it does not need to measure the absolute distance accurately in order to get an accurate speed reading.

But, as it turns out, crystal oscillators are so accurate that any errors would be slight anyway, even if not compensated for. Unless of course something went wrong with the electronics, in which case the clock may be wildly wrong - but such devices are normally equipped with watchdog circuits which detect when the clock is out of spec.



Edited by Pete317 on Wednesday 25th November 20:43
Not quite. Such independence in clock would require the speed of light to be dependent on the clock too.


Edited by V8LM on Wednesday 25th November 21:42
You're missing something - or maybe I'm not explaining it as clearly as I should.
A shortness in the perceived delay is exactly compensated for by the longer actual time between consecutive pulses, which makes the perceived difference between the two delays correct.


Edited by Pete317 on Wednesday 25th November 21:57
True. And apologies to tapereel.

ETA:

first distance = clicks1 * time_per_click * c / 2
second distance = click2 * time_per_click * c / 2
time between clicks = click12 * time_per_click

velocity = (clicks2 * time_per_click * c / 2 - click1 * time_per_click * c / 2) / clicks12 * time_per_click

velocity = (click2 - clicks1) / clicks12 * c / 2

Edited by V8LM on Wednesday 25th November 22:33

Pete317

1,430 posts

221 months

Wednesday 25th November 2015
quotequote all
V8LM said:
Pete317 said:
V8LM said:
Pete317 said:
Actually, he's right in the bit about the speed being right, except that the other things he says leads me to believe that he doesn't know why he's right, and is probably just parroting what someone has told him.

If, for example, the timing clock is running slow, then it will measure the pulse propagation time as less than it actually is.
But, because the rate at which the pulses are sent is controlled from the same clock, the pulses will also be sent at a slower rate.
This means that the inaccuracy in the measured pulse propagation time will be canceled out by the fact that the time between the pulses will also be longer, and so the difference between the two pulse delays will work out the same, regardless of the inaccuracy of the clock.

So yes, an inaccuracy of the clock does not affect the speed reading, but it doesn't alter the fact that the device works by measuring distance and converting the relative distance into speed. The fact that it may not measure the distance accurately is immaterial, it still measures the distance but it does not need to measure the absolute distance accurately in order to get an accurate speed reading.

But, as it turns out, crystal oscillators are so accurate that any errors would be slight anyway, even if not compensated for. Unless of course something went wrong with the electronics, in which case the clock may be wildly wrong - but such devices are normally equipped with watchdog circuits which detect when the clock is out of spec.



Edited by Pete317 on Wednesday 25th November 20:43
Not quite. Such independence in clock would require the speed of light to be dependent on the clock too.


Edited by V8LM on Wednesday 25th November 21:42
You're missing something - or maybe I'm not explaining it as clearly as I should.
A shortness in the perceived delay is exactly compensated for by the longer actual time between consecutive pulses, which makes the perceived difference between the two delays correct.


Edited by Pete317 on Wednesday 25th November 21:57
True. And apologies to tapereel.
He may have been right on that little point, but he's not going to get any apologies from me for wrongly accusing others of being wrong.

Pete317

1,430 posts

221 months

Wednesday 25th November 2015
quotequote all
V8LM said:
first distance = clicks1 * time_per_click * c / 2
second distance = click2 * time_per_click * c / 2
time between clicks = click12 * time_per_click

velocity = (clicks2 * time_per_click * c / 2 - click1 * time_per_click * c / 2) / clicks12 * time_per_click

velocity = (click2 - clicks1) / clicks12 * c / 2
You've just explained it more clearly than I could

V8LM

5,166 posts

208 months

Wednesday 25th November 2015
quotequote all
In the not-too-sober darkness of night, it's obvious: One can measure a velocity using something with a known velocity (providing you have a stable clock), but to measure a distance using something with a known velocity one needs to measure time too (or to measure time one needs a known distance).


Edited by V8LM on Thursday 26th November 07:21

Monty Python

4,812 posts

196 months

Thursday 26th November 2015
quotequote all
I think claim 1 in the patent sums it up nicely:

"said control means being configured to:
calculate a plurality of distances to said target from said flight times and the speed of light, each corresponding to a different successive ones of a series of laser pulses; and calculate a velocity of said target relative to said observer based on said distances and the elapsed time between the individual said pulses corresponding to said distances."

tapereel

1,860 posts

115 months

Thursday 26th November 2015
quotequote all
V8LM said:
In the not-too-sober darkness of night, it's obvious: One can measure a velocity using something with a known velocity (providing you have a stable clock), but to measure a distance using something with a known velocity one needs to measure time too (or to measure time one needs a known distance).
...and Hey Presto! A speed measurement as a ratio of the speed of light with no distance in feet, metres, inches, miles required or computed.
The speed of light doesn't change significantly in respect to the speed calculation anywhere on earth so the device is always calibrated, hence a self-calibrating device. I don't think you can get a better standard for calibration than a fundamental physical property.

V8LM said:
I think claim 1 in the patent sums it up nicely:

"said control means being configured to:
calculate a plurality of distances to said target from said flight times and the speed of light, each corresponding to a different successive ones of a series of laser pulses; and calculate a velocity of said target relative to said observer based on said distances and the elapsed time between the individual said pulses corresponding to said distances."
Well yes it does but have you realized that those distances are measured in units of time that are essentially arbitrary. The clock "clicks" as V8LM puts it.
You may also see in the patent that the distance indication in the speedmeter is computed separately to the speed.

Well done everyone, now you know I was right. smile

pip-pip

tapereel

1,860 posts

115 months

Thursday 26th November 2015
quotequote all
Pete317 said:
He may have been right on that little point, but he's not going to get any apologies from me for wrongly accusing others of being wrong.
...even though they were.

No apologies required; your subsequent education and realization is all of the satisfaction I need.

Rovinghawk

13,300 posts

157 months

Thursday 26th November 2015
quotequote all
tapereel said:
...and Hey Presto! A speed measurement as a ratio of the speed of light with no distance in feet, metres, inches, miles required or computed.
Quoted so you can't edit anything when you realise it's bks.

You can't compute speed without a distance between 2 points & a time interval between those points. I accept that this distance can be computed by time delay before reflection of a pulse but distance nevertheless needs to be ascertained as a stage towards speed calculation.

My degree's in engineering, with rather a strong bias on this sort of stuff. Do you have any form of education?

Monty Python

4,812 posts

196 months

Thursday 26th November 2015
quotequote all
Rovinghawk said:
Quoted so you can't edit anything when you realise it's bks.

You can't compute speed without a distance between 2 points & a time interval between those points. I accept that this distance can be computed by time delay before reflection of a pulse but distance nevertheless needs to be ascertained as a stage towards speed calculation.

My degree's in engineering, with rather a strong bias on this sort of stuff. Do you have any form of education?
I'm at a loss for words....

You cannot measure the speed of car using a hand-held device such as this without knowing both distance and time - otherwise why does it bother taking a series of measurements rather than just one?

"These LIDAR devices typically send out a stream of 100 pulses within three-tenths (0.30) of a second. By measuring the length of time for each pulse to be sent and returned and dividing that number by two, it can calculate the distance to the vehicle. By comparing the distance calculated by each subsequent pulse, the speed can be determined."

The only instance where you can measure the speed of an object with one measurement is if the object itself is emitting light and you measure the red/blue shift. This technique can measure down to 10mph but the equipment is by using a spectrometer.

If this is wrong, then please explain to me how you can calculate the speed of a car from one measurement.




Edited by Monty Python on Thursday 26th November 10:55

Pete317

1,430 posts

221 months

Thursday 26th November 2015
quotequote all
tapereel said:
V8LM said:
In the not-too-sober darkness of night, it's obvious: One can measure a velocity using something with a known velocity (providing you have a stable clock), but to measure a distance using something with a known velocity one needs to measure time too (or to measure time one needs a known distance).
...and Hey Presto! A speed measurement as a ratio of the speed of light with no distance in feet, metres, inches, miles required or computed.
The speed of light doesn't change significantly in respect to the speed calculation anywhere on earth so the device is always calibrated, hence a self-calibrating device. I don't think you can get a better standard for calibration than a fundamental physical property.
Bullst! It's got nothing to do with the speed of light.

You have to obtain some form of distance measurements, whether expressible in units of distance or not, in order to calculate the speed.
Other than using the Doppler effect, which can't be done in any practical sense with laser devices, there's no mechanism by which you can obtain speed without first obtaining some measure of relative distance over a time period. Just because you can't necessarily express the distance measurements in units of distance doesn't alter that fact.

Edited by Pete317 on Thursday 26th November 11:44

tapereel

1,860 posts

115 months

Thursday 26th November 2015
quotequote all
You chaps are all missing the point that is in contention.

The distance displayed on laser speedometers need not be and is not that used in the speed calculation. Because of the good design and quality of the instruments the displayed speed is usually, by not always, consistent to the distance around the mid point of the measurement data.

The method of using the speed of light as a standard, the ratio of which is used to indicate target speed, creates the ability of devices to operate without the absolute rate of oscillation of the system clock remaining as a critical parameter in the speed calculation. The only critical property of the system clock is that it remains at the same frequency for all samples taken during the speed measurement. It is trivial to get an instrument to monitor that stability.

The ratio of the target speed to the speed of light is determined by making a series of delta range measurements in units of time. Those units of time are a function of the clock period which is altered to increase it resolution. The clock frequency or time period is not used to resolve the distance to target or delta distance between samples to any standard unit of distance in the speed measurement so the clock frequency is immaterial except for its stability.

Say the clock frequency is exactly as designed, after he speed is resolved a calculation is made, using the assumed clock period to indicate target distance in feet or metres.

Now let's look at ratios. If the correct clock period gives a ratio of (n2-n1)/ni. Or (40-20)/20 = 1

Now half the clock rate (20-10)/10 =1

Double it. (80-40)/40 =1

As the ratio is the multiplier to resolve the speed of the target from the speed of light this target is traveling at the speed of light. Not a practical target but I figure simple numbers are best.

When you take the clock period in seconds to multiply the speed of light to resolve the distance to the target the assumed clock frequency is critical or your distance will change proportional to the clock period. You can see however that the clock period assumed or otherwise is not used to convert to standard units of distance because the method of determining the ratio of target speed to C makes the distance in standard units superfluous and damaging to a self calibrating speed measuring device.

I too have an engineering degree and more than 40 years of engineering practice in which I have excelled. What that experience and teaching has meant though is I am aware that high educational qualifications don't mean you know everything in that discipline. The sooner others realise that the more they will be able to gain from their learning.

Pip pip.


Monty Python

4,812 posts

196 months

Thursday 26th November 2015
quotequote all
But you still need two measurements of the distance to the target in order to determine the speed - you can't do it from one measurement.

You're correct that the speed of light is involved, but only to calculate the time the laser pulse takes to hit the target and return to the gun. You still need more than one distance measurement to determine the speed.

"The algorithms processed by a microprocessor of a LiDAR gun works on a basic definition of speed i.e. speed is equal to ratio of distance travelled by and time taken to cover the distance. Initially a pulse(X) is released from the gun and gets reflected back to the gun on striking the target after a time interval (t1). The time interval (t1) is halved, as t1 is the time taken for the full flight. The time interval (t1) is multiplied with velocity of light as velocity of infrared waves is same as light velocity. From the above calculations we can determine the distance of the target at the second instant. We can find the velocity of the target by calculating the ratio of the distance travelled from first instant to second instant and difference in the time intervals for the first and second instant. In reality, the above calculations are performed by many impulses and then the velocity of the target is displayed on the LCD. We can also know whether the target is accelerating or decelerating."

(from Proceedings of 5th IACEECE, 2013).

http://www.speed-trap.co.uk/Accused_Home/How_it_Wo...

"LIDAR devices typically pulse at between 45Hz and 72Hz - i.e. 45 to 72 times a second. The speed of light is 300million metres/second. Let's take two pulses from a LIDAR device as an example.
Pulse 1 emitted from the device and detected by the receiver 0.000001333 seconds later. Well, speed equals distance over time. So distance = speed x time. The speed of light is 300,000,000m/s and the time is 0.00000133s so the distance is
300,000,000 x 0.000001333 = 399metres. That's the distance the light travelled from the gun to the target and back again, so the distance to the object is half that, or in this case 199.5metres
Pulse 2 is emitted 1/45th of a second later, and detected 0.000001325 seconds after that. The same calculation again gives the distance of
300,000,000 x 0.000001325 = 397.5metres round-trip, or a device-to-target distance of 198.75metres.
Now it's another calculation using the above figures. The difference in distance between the two readings is 0.75metres. The time between pulses was 1/45th of a second so the object travelled 0.75 metres in that time, or
0.75 x 45 = 33.75metres/second. It's now even simpler maths to scale that up from metres/second to km/h -
33.75m/s x 3600 = 121500metres in an hour which in turn is 121.5km/h. Now bear in mind that all that those calculations happen 45 times a second (or more). The device needs two or three consecutive readings all with similar results to give a finite speed which is why LIDAR devices take 0.3 seconds to lock-on and read your speed."

Again, two measurements of distance are required to calculate the speed - there's no getting away from it.

Rovinghawk

13,300 posts

157 months

Thursday 26th November 2015
quotequote all
tapereel said:
I too have an engineering degree and more than 40 years of engineering practice in which I have excelled.
So why are you talking crap now?

Distance is relevant to measuring speed. To deny this is to talk bks.

Getting back to the original point, optronic measurement devices have a margin for error despite any claims to the contrary. They are only good for the conditions at which they are calibrated and even then not necessarily for prolonged periods of time.

tapereel

1,860 posts

115 months

Thursday 26th November 2015
quotequote all
Monty Python said:
But you still need two measurements of the distance to the target in order to determine the speed - you can't do it from one measurement.

You're correct that the speed of light is involved, but only to calculate the time the laser pulse takes to hit the target and return to the gun. You still need more than one distance measurement to determine the speed.

"The algorithms processed by a microprocessor of a LiDAR gun works on a basic definition of speed i.e. speed is equal to ratio of distance travelled by and time taken to cover the distance. Initially a pulse(X) is released from the gun and gets reflected back to the gun on striking the target after a time interval (t1). The time interval (t1) is halved, as t1 is the time taken for the full flight. The time interval (t1) is multiplied with velocity of light as velocity of infrared waves is same as light velocity. From the above calculations we can determine the distance of the target at the second instant. We can find the velocity of the target by calculating the ratio of the distance travelled from first instant to second instant and difference in the time intervals for the first and second instant. In reality, the above calculations are performed by many impulses and then the velocity of the target is displayed on the LCD. We can also know whether the target is accelerating or decelerating."

(from Proceedings of 5th IACEECE, 2013).

http://www.speed-trap.co.uk/Accused_Home/How_it_Wo...

"LIDAR devices typically pulse at between 45Hz and 72Hz - i.e. 45 to 72 times a second. The speed of light is 300million metres/second. Let's take two pulses from a LIDAR device as an example.
Pulse 1 emitted from the device and detected by the receiver 0.000001333 seconds later. Well, speed equals distance over time. So distance = speed x time. The speed of light is 300,000,000m/s and the time is 0.00000133s so the distance is
300,000,000 x 0.000001333 = 399metres. That's the distance the light travelled from the gun to the target and back again, so the distance to the object is half that, or in this case 199.5metres
Pulse 2 is emitted 1/45th of a second later, and detected 0.000001325 seconds after that. The same calculation again gives the distance of
300,000,000 x 0.000001325 = 397.5metres round-trip, or a device-to-target distance of 198.75metres.
Now it's another calculation using the above figures. The difference in distance between the two readings is 0.75metres. The time between pulses was 1/45th of a second so the object travelled 0.75 metres in that time, or
0.75 x 45 = 33.75metres/second. It's now even simpler maths to scale that up from metres/second to km/h -
33.75m/s x 3600 = 121500metres in an hour which in turn is 121.5km/h. Now bear in mind that all that those calculations happen 45 times a second (or more). The device needs two or three consecutive readings all with similar results to give a finite speed which is why LIDAR devices take 0.3 seconds to lock-on and read your speed."

Again, two measurements of distance are required to calculate the speed - there's no getting away from it.
Who says you don't?.

Approved systems use between 40 and 60.

tapereel

1,860 posts

115 months

Thursday 26th November 2015
quotequote all
Rovinghawk said:
tapereel said:
I too have an engineering degree and more than 40 years of engineering practice in which I have excelled.
So why are you talking crap now?

Distance is relevant to measuring speed. To deny this is to talk bks.

Getting back to the original point, optronic measurement devices have a margin for error despite any claims to the contrary. They are only good for the conditions at which they are calibrated and even then not necessarily for prolonged periods of time.
Reasoned discourse is wasted upon you.