Detecting a bullet for velocity calculation

[blackflag267x]

Trying to detect voltage differences, probably on the magnitude of nanovolts or possibly picovolts. I don't need to read the actual voltage, just detect the occurence of a change. The goal is to detect changes induced in an coil around a magnet... any insight on that would be helpful too.

Pretty new to most circuitry, but from what I've read a comparator or open-loop op amp (?) may do the trick. I'll ultimately be using arduino uno to do the detection.

Thanks!

 

[Arouse1973]

You will need some special circuitry for that. Do you know how small that is? To put it in context a 2K resistor at 20deg C will produce about 800nV of noise when not even connected to anything. How have you worked out it might be that small?
Adam

 

[Gryd3]

As Adam has stated, that is an incredibly small value to attempt to measure one billionth of a volt. You would need to multiply it by a million just to get a millivolt.
http://en.wikipedia.org/wiki/Nano-
This would be a lab grade test requiring shielding to prevent outside interference/noise. Is this a personal project?

 

[davenn]

The goal is to detect changes induced in an coil around a magnet... any insight on that would be helpful too.

voltages are likely to be much larger than you suspect/suggest

Again ... what is your project ... in detail ... so we don't have to play 20 questions trying to guess an answer for you

Dave

 

[blackflag267x]

Thanks for the responses. In retrospect, yes, that is a very small voltage; it probably isn't that small. It's a personal project and unfortunately, my name isn't Dexter.

Trying to sense when a copper bullet at ~2700 fps moves over the magnet+coil. Theoretically, the moving metal should momentarily change the magnetic field and induce or change the current in the coil right? (Lenz's law?) I was just messing around with some cheap magnets and 30 gauge wire and got a reading of about 1 uA with an ammeter when I ran two magnets (one with probably 200 coils) past eachother. I figured if the bullet was inducing a current, it would be very small. Since then, I've been parusing the internet and trying to understand comparators, op amps, input offset voltages, etc, but am not sure what exactly to look for or if I'm on the right path. Unfortunately my Structural engineering degree doesn't quite translate into an EE.

I know there's a way to detect the bullet because there are such commercial products on the market. But building something yourself is much more satisfying (and budgetary) than buying it so that's the approach I'm trying to take.

Thanks again for the help!

 

[Gryd3]

Thanks for the responses. In retrospect, yes, that is a very small voltage; it probably isn't that small. It's a personal project and unfortunately, my name isn't Dexter.

Trying to sense when a copper bullet at ~2700 fps moves over the magnet+coil. Theoretically, the moving metal should momentarily change the magnetic field and induce or change the current in the coil right? (Lenz's law?) I was just messing around with some cheap magnets and 30 gauge wire and got a reading of about 1 uA with an ammeter when I ran two magnets (one with probably 200 coils) past eachother. I figured if the bullet was inducing a current, it would be very small. Since then, I've been parusing the internet and trying to understand comparators, op amps, input offset voltages, etc, but am not sure what exactly to look for or if I'm on the right path. Unfortunately my Structural engineering degree doesn't quite translate into an EE.

I know there's a way to detect the bullet because there are such commercial products on the market. But building something yourself is much more satisfying (and budgetary) than buying it so that's the approach I'm trying to take.

Thanks again for the help!

I wonder if the bullet will be too fast to induce a current for a long enough period to detect or manipulate.
Have you considered an optical solution?

 

[blackflag267x]

Wouldn't the ability to detect a simple change or spike in current be dependent on the sampling rate of the sensor? So if you could get a part capable of taking microsecond readings, it would be able to sense a change? At 2700 fps the bullet would travel through a 1" distance in 31 microseconds. Because it has been done in a commercially available product, I thought it might be doable by a hobbyist as well.

I have considered the optical approach. I originally thought the electromagnetic solution might be easier from the DIY approach and less prone to sampling error (detecting light variations would be pretty dependent on ambient conditions). Maybe using store bought lasers and some simple photodiodes would reduce this problem, but they would have to exactly in line with the bore.

 

[Gryd3]

The 'sampling rate' does not really apply unless you want a microcontroller to read the values. If you build a linear circuit, it can read and react extremely quickly. What you need to consider though is how quickly the circuit will react. The 31μs pulse you have mentioned loosely translates to a single pulse from a 33KHz clocked device which is definitely detectable by building a circuit to detect the rising edge of this pulse.
There may be complications depending on the parts chosen for the final circuit though. You need to make sure that the pulse that is picked up can be amplified or buffered sufficiently to prevent it from degrading or getting hidden within the noise picked up by external forces.

Will your device be a large loop that the bullet passes through, or a small disk that the bullet flies over?
If you have access to an oscilloscope, you could experiment to determine an ideal inductor to build to detect the bullet.

 

[blackflag267x]

I was planning on using a microcontroller - two "switches" sitting x distance apart have to be tripped and the time between their signals measured. Trying to measure velocity, I guess I left that out earlier. I figured an arduino uno has ample clock speed at 16 MHz; though maybe there are some intricacies I'm missing.

The signal amplification is what I originally posted for. I'm not sure how to even begin detecting small voltages like this.
I was thinking I'd use a magnet with wire coiled around it and orient the centerline of the coil perpendicular to the bullet trajectory. I figured this was similar to the way the coils in a metal detector would be oriented to the metal they are detecting, though the way in which they detect is different. It's also similar to the way guitar pickups are set up and oriented, though the direction of movement is different. It seems to me that if the bullet were to pass through the center of a loop (with no magnet and an AC current) it would have less effect on the field although that is just unfounded conjecture. I don't have access to an oscilloscope but may know people who do.

 

[Gryd3]

If you can get an oscilloscope that can 'record' make a couple coils and field test it.
Hopefully you can get a coil that will respond quick enough to produce a voltage that you can manipulate.
There are devices you can use to do so. It sounds as though you don't need to 'measure' it so much as to detect it's there. This makes it easier. We still need to determine if what you make is good enough to pick up the bullet
What your building though may be difficult for us to help.
I don't have a gun to test things with, and my spit balls only go so fast...

 

[blackflag267x]

Yes, detection is all I need. Would the response time vary with the reactance (and thus the inductance) of the coil? So to a achieve a quick reaction time with a given coil resistance and frequency, I'd need a low inductance coil, though I'm not quite sure what this means or how to achieve it. Having a magnet inside the coil would heavily increase the reactance of the coil so using an oscillating current to produce a small magnetic field be better. Probably won't get access to an oscilloscope, unless I can find one on craigslist or something.

Based on some further reading, I believe the sensor setup I need to pursue is similar to industrial inductive proximity sensors. Thus I'd need an oscillator, a Schmitt trigger with a high slew rate and (maybe?) a demodulator. Still just spitballing, apologies if what I'm saying makes no sense

 

[Gryd3]

It makes sense to me, but I have no experience with that type of circuit. I can spitball with you using some of the info I picked up in college though.

 

[blackflag267x]

Ok, thanks for the info. Keeping everything else the same, would increasing the number of coils in a loop increase the strength of the magnetic field? And would an outside object (in this case the bullet) have more effect on the field if the original field is weaker or stronger? Or maybe there's a sweet spot that induces enough current in the bullet to cause a detectable change.

 

[Gryd3]

Ok, thanks for the info. Keeping everything else the same, would increasing the number of coils in a loop increase the strength of the magnetic field? And would an outside object (in this case the bullet) have more effect on the field if the original field is weaker or stronger? Or maybe there's a sweet spot that induces enough current in the bullet to cause a detectable change.

This is about the extent of my knowledge :S .. I know that adding coils will add inductance, but I am unsure how that translates to sensitivity. I also know that more coils will allow a larger voltage to develop when a magnetic field changes (but current goes down... we only need to detect the voltage though)

 

[Arouse1973]

You basically want a fast metal detector. As the bullet passes by it will detune the metal detector by virtue of the induced eddy currents into the copper bullet. This is going to be small and very fast. I have designed a couple of metal detectors in the past and there are a couple of different approaches open to you.

Although I never tried it with a passing bullet so the latency of the detector might be too great. First option is to create a high Q tuned circuit and detect the amplitude reduction of the LC circuit when it is detuned by the bullet. Second is to have a lower Q circuit that is not detuned much in the presents of the bullet and detect the increase in current.

Third is to excite a tuned circuit with a pulse of energy and let it ring then you measure the degreasing amplitude of oscillations at a given point you then you do the same and measure again and if metal is present, the amplitude of oscillations decay sooner. But with the speed of the bullet it might be too fast I don't know.

Adam