Accurate Low Speed Measuring

just a good old fashioned radar gun is not good enough?

The typical radar gun even those used for sports at best will measure speeds down to 6 mph, less than that they wont work plus their error is around ± 1 mph.

I did fail to mention that the speeds I am working with range between 1-8 mph.

You should be able to put together an ultrasound unit for about $200. Or, to be really creative, look at Sensors (used to be magazine, now website)

http://www.sensorsmag.com/content/search?keywords=velocity

If the measurement technique is not at issue, you could just repeat the configuration of 2 sensors + PLC, using sensors (through-beam or retro-reflective optos?) from a different manufacturer, and a simple logic module (Moeller Easy, Siemens LOGO!, Crouzet Millennium etc) with a built-in display.

Use as big a separation between the sensors as possible/practical.

"Use as big a separation between the sensors as possible/practical."

I would love to have at least 35 inches or more for calculating the speed but unfortunately the sled along with the rest of the system has to fit in a building that has already been built so the available footprint is smaller that any of us would like. To reach our top sled speed of 8 mph most of the available sled travel is taken up just accelerating the sled to speed which leaves 24 inches at best to measure the the final sled speed. The next thing some may suggest is use a larger motor to accelerate the sled faster but the mass of the motor becomes an issue when trying to stop the motor, the drive chains and pusher blocks.

As for using a second dedicated PLC and a different set of sensors to repeat the measurement, it is a very good idea, one which I may need to do to keep costs down. The only thing is it will require "Operators" to be very accurate about setting the spacing of the sensors if they move the "Test Unit" from sled to sled. I would first llike to find a method/instrument that is plug and play as much as possible.

For sensor spacing - a fixture (2" x ¼" (50mm x 6mm) ally bar or whatever with the sensors mounted thereon) should eliminate errors in test unit installation.

If you have access to a digital storage scope, you might set up two photocells and two laser pointers spaced an appropriate distance apart (1 ft, or...?) so that the laser beams cross the sled's track. The photocell signals are fed into 2 channels of the o'scope. As the sled breaks the beams, the differential loss of signals from the two photocells can be read from the scope trace as time. The distance and time will of course give you the speed. You might need the services of an EE if the photocell signals need to be amplified, but that could be kludged fairly easily, I think.

GA Usbport- the o'scope will give you very accurate data that does not require any sort of time delay for digitizing the signal. I should think there would be no need to amplify the signal from the photo detectors if one is using a bright enough source light (i.e., a laser). The accuracy of the measurement would depend primarily on how accurately you know the distance between the sensors...

Nice idea but not user friendly. The people needing to use it are lucky to have graduated high school......if they even did.

Too simple. In the "good old days" use a high speed clock (digital or sweep hand) and videotape (Showing my age) the passage of the sled.

Review of the recording will provide you with an accurate transition time and then it's a matter of distance/time to get your velocity.

Absolutely non contact, absolutely repeatable. Only dependant on your ability to read the clock.

In the modern world, digital cameras have inbuilt clocks that can be displayed in the recorded signal. The camera is then available for other use in the lab.

The drawback is that it doesn't directly give you a "speed" that can be electronically interrogated but since this seems to be used to proove your current set-up it might be adequate.

Using a camera is a possibility to follow. The customer already is using vision systems as part of their inspection section so they should be able to configure a camera for speed measurement. I will have to research the cost of this option. It just might be a cost effective option.

You want a second means to double the chance of it going wrong?
The big question is WHY?

There are a myriad of simple solutions, but it really depends what you ACTUALLY need.
You could simply duplicate the existing system.
Or...
It is very simple to build a circuit.
The first sensor triggers two timers...
The second sensor must be triggered after the first timer has timed out and before the second has timed out.
It's a basic bit of electronics and logic.
Failure to trigger the second sensor in the correct time can sound an alarm illuminate a big flashing light, unleash a spring loaded boxing glove and automatically open a tin of Tuna for any cats in the vicinty.

The circuit can be reset by a short delay after the second sensor or by umpteen different methods.
It doesn't even need a microcontroller.
A good electronics hobbyist (or a cat a squirrel or maybe a sack of Gophers?) should be able to do it.
Del

Good suggestion, but I know my customer and they would want something a little less fishy, perhaps can chicken instead of tuna.

Can you use a high speed video capture system to verify velocity?

In many impact studies, HS video provides important dynamic information that may be lost by only doing a postmortem analysis. If a new camera is too expensive for you to purchase, demo cameras and rentals are usually available at significantly lower prices.

You can use an LVT (Linear Velocity Transducer). See this URL:

http://www.transtekinc.com/products/linear-velocity-transducers.html

You may have to mount a magnet on the sled.

Nice idea but it requires physical contact and is limitted to 24 inches of usable travel, my sled as more travel than that. I would aslo be concered with shock damage since my sled will be slamming into fairly solid objects and coming to abrupt stops several times during a typical day of testing..

What is not defined is the sled speed and resolution you seek.

Del's idea of a counter started and stopped by the sled passage between two points.

Calculate the counter's frequency to get proper resolution time/speed and the number of bits for the counter's elapsed time.

Optical triggering is best, beware of industrial sensors which may have internal filtering and added (temperature variable) time constant.

I did miss the speed range of the sled, it is 1-8 mph. The accuracy needs to be 2% of reading or better.

Sorry for the delay in responding, I have been on the road for the past week with less than stellar internet connections.

This is not a school project.

Now that I am back in the office I will follow the links provided and review eveyones suggestions. Thank you to those that have responed so far.

These are not accurate at low speeds and require a lot more time to respond than I have available in this sled.