Alternative to Relay for Accurate Time Measurement Experiments

[Harsh Bhardwaj]

I am carrying out a physics experiment to measure g by free fall using an Arduino. This experiment requires accurate time measurements of upto a millisecond. I am using a relay to turn on/off an electromagnet and there is a maximum lag of 10ms due to the relay. I am a complete beginner. Please suggest alternative mechanisms to switch the electromagnet and save those precious milliseconds. P.S. - the magnet is driven by 12V 2.5A DC.

UPDATE

I tried something like this using an NPN 2N2222A transistor. This transistor couldn't hold the required current for the electromagnet and a lower power supply was not strong enough.

 

[mfb]

You can get a better transistor. A few Volt, a few Ampere - should cost less than shipping costs if buying one is an option.

An alternative on the physics side: Measure the time at two places, or measure the time as function of height in several steps, and determine g and the relay time delay both from experimental data.

 

[Nidum]

Problem may not all be due to the relay response .

The electromagnet itself is not going to turn on and off instantaneously .

If electromagnet response time does turn out to be a problem then you can improve it considerably using a higher voltage drive and some slightly more complicated electronic switching .

 

[Harsh Bhardwaj]

Problem may not all be due to the relay response .

The electromagnet itself is not going to turn on and off instantaneously .

If electromagnet response time does turn out to be a problem then you can improve it considerably using a higher voltage drive and some slightly more complicated electronic switching .

As you warned, the magnet is playing a huge role. I am getting g as about 860 cm/s. I believe this lower value is because initially there is an upward acceleration due to the magnet on the falling object. I am using a DIY electromagnet with an iron bolt as its core. How can I make the magnetic force die away instantly?
To make the relay lag error less significant I increased the fall height to about 200 cm, thus increasing the fall time. I plan to buy transistors that can take the 12V 1.5A DC load.
If the magnet fails then I will look into using mechanical platform driven by a motor to make the object fall suddenly.

 

[Harsh Bhardwaj]

UPDATE:
I now used a heavier mass and it gave me values closer to the accepted value of g. I think this is because the gravitational force(proportional to mass) now dominates over the magnetic force. I am surprisingly getting about 950 to 960 cm/s. I will upload the Arduino code if the need be.

 

[Nidum]

If the magnet fails then I will look into using mechanical platform driven by a motor to make the object fall suddenly.

Something like that would be much better - you are never going to get the response time of the solenoid down to nothing .

A common set up is for the test object to be held in place by a latch - something like a door bolt but a bit more sophisticated . The solenoid then just pulls the bolt out of engagement - the actual point of release is detected by an independent detection system .

Something to always take into account with latch and release systems is to make sure that they don't put any impetus into the test object or launch it slightly crooked . For this reason latch and release systems are often arranged to be axially symmetric . Most common arrangement is to have three latches arranged radially . That's just for interest - you'll do perfectly ok with a single latch if you are careful with it's design .

 

[Nidum]

I now used a heavier mass and it gave me values closer to the accepted value of g. I think this is because the gravitational force(proportional to mass) now dominates over the magnetic force. I am surprisingly getting about 950 to 960 cm/s. I will upload the Arduino code if the need be.

That's a good result . Entirely logical since the heavier object will overcome any residual solenoid force earlier on in the transient turn off period .

 

[jim hardy]

Please suggest alternative mechanisms to switch the electromagnet...

I think geophysicists use a pendulum to measure local gravity. That gets you past any starting transients.

https://spacemath.gsfc.nasa.gov/weekly/10Page39.pdf

 

[Harsh Bhardwaj]

I think geophysicists use a pendulum to measure local gravity. That gets you past any starting transients.

https://spacemath.gsfc.nasa.gov/weekly/10Page39.pdf

I just completed my first semester in undergrad physics. We were supposed to do many experiments to determine g like Kater's pendulum, spring oscillations, bar pendulum etc. This particular experiment - using digital timing techniques was omitted. So I decided to give it a shot in these vacations. I am just trying it in my free time.

 

[jim hardy]

using digital timing techniques was omitted. So I decided to give it a shot in these vacations. I am just trying it in my free time.

Sounds like fun -you just have that release transient to overcome.

 

[CWatters]

An alternative on the physics side: Measure the time at two places, or measure the time as function of height in several steps, and determine g and the relay time delay both from experimental data.

+1

Google the ball falls past a window problem. Typically an observer sees a ball falls past a window of known height taking a known time and they ask how high above the window the ball was dropped. If you know that you can instead solve for g eliminating the relay time delay from the calculation.

 

[vk6kro]

A quick release can be made by reducing the voltage on the electromagnet so it is only just holding the falling object.

For the falling object, do not use just a piece of iron.
This will magnetise the iron and it will take a short while to let go.
A better falling object is a ball from an old rolling ball computer mouse. These are rubber coated iron and let go of the electromagnet instantly.
And it does less damage to anything it hits.

When the ball lands, you can arrange a small seesaw. The ball lands on one side which is held up by a spring.
The other side pushes down on a microswitch.
The ball falls pushing the seesaw down which lifts the other side up releasing pressure on the micro switch.

Note that you may need a normally closed micro switch to do this.

 

[sophiecentaur]

Please suggest alternative mechanisms to switch the electromagnet and save those precious milliseconds.

There is no need to involve the release time if you use three or four measurement points as the mass falls past them. That way allows you to measure the distance / time relationship over multiple points which gives you a good value for g, whenever the ball was released. Plot the (d,t) points on a graph and fit a parabola to them. Only one parabola will fit them and that one has the right value for g.
If the measurement systems are identical then so will be their reaction time - which means that this common error will be eliminated.

 

[vk6kro]

Your electromagnet is probably too powerful.
You could use a relay coil with the contacts removed.

Place a computer mouse ball at the end of the relay coil and apply just enough voltage to attract the ball.

You can drive this relay coil with a small power transistor. Just remove the power to release the ball.
This is easy with an Arduino.

This experiment is usually done to work out the acceleration due to gravity.
This varies slightly with location, but is well known, so you can use the known value to confirm that your equipment is working OK. It should be about 9.81 metres/sec/sec.

Set the drop to exactly one metre and measure the drop time. If this corresponds with the calculated value, your apparatus is working OK.

 

[sophiecentaur]

@https://www.physicsforums.com/threads/alternative-to-a-relay.935556/members/harsh-bhardwaj.622914/j
You will never get your best possible answer if you are logging the time differently at each end of the journey of the falling mass. Have you done any significant amount of research about this? If you google terms like "measuring g" and "students" you will get loads of information. You need at least three identically measured times on the drop. All other approaches will just be attempts to reduce an error that needn't be there in the first place.
Having said that, there is school level equipment with mechanical contacts and a simple electromechanical timer that will give an 'acceptable' result. That's what they used before fancy electronics was available in school labs.

 

[Harsh Bhardwaj]

I searched about this experiment on Google before beginning with it myself. Most of these got a value correct to one significant figure only and I got that too. To get greater accuracy I'll follow the 3 points approach this week. My professor suggested that I can study the nature and distribution of random errors in this experiment by taking large number of measurements. Originally I studied random error by measuring dimensions of hundreds of kidney beans. Thank you all for the help. I'll be reporting about any progress.

 

[vk6kro]

The error with this experiment comes from using a steel ball bearing as the falling object.

I used an electronic timer with 4 decimal place accuracy. The only variation in results was in the last figure and then only a slight variation.

But the secret was to use the ball from an old computer mouse. This completely cut out the obvious delay where the ball would not fall when released.

 

[vk6kro]

Here is a useful website for calculating the time for an object to fall a known distance.
http://keisan.casio.com/exec/system/1224835316

It just uses the standard formula, but it is a handy check.

Your local observatory should be able to give you an accurate value for "g", the acceleration due to gravity, at your location. Mine is 9.79 m/sec^2, so it does vary.

For a 1 metre fall, the time should be about 0.45 seconds.

If you are using an Arduino to do your measurement, you can get better accuracy by using interrupts.