Coilgun projectile detection
Brendon, Mon Feb 12 2007, 09:58PMHi,
I've come up with an idea for detecting the position of a projectile with capacitance. I've included a drawing to show what I mean. Has this been done before and can it work?
Here is how I think it should work.
1. The projectile goes between the two plates increasing the capacitance.
2. Then the projectile leaves the capacitor can't dischange through the diode, so the voltage across the opamp should increase and trigger it.
3. The capacitor dischanges through a large resistor ready for the next shot.
The capacitance with the bullet in the center should be about 2pF - 10pF (larger for longer and wider bullets) and the capacitance across the diode and opamp together would be about 6pF so the increase in voltage should be measurable.
Brendon
[Edit: Picture size]
Re: Coilgun projectile detection
TheMerovingian, Mon Feb 12 2007, 10:12PM
Theoretically iy would work if the magnetic field variation doesn't interfere with the detection circuit. I think that photogates are way simpler
TheMerovingian, Mon Feb 12 2007, 10:12PM
Theoretically iy would work if the magnetic field variation doesn't interfere with the detection circuit. I think that photogates are way simpler
Re: Coilgun projectile detection
Simon, Mon Feb 12 2007, 11:11PM
This might work well if typical CG projectiles had high permittivity but they don't. CG projectiles (for this kind of CG) are designed for high permeability - i.e. to be pulled by a magnetic field, not an electric field.
Besides, you've got a big magnetic pulse going. This is likely to swamp any very small signals.
It's a nice idea for being different but I think I'd stick with photogates.
Simon, Mon Feb 12 2007, 11:11PM
Brendon wrote ...
The capacitance with the bullet in the center should be about 2pF - 10pF (larger for longer and wider bullets) and the capacitance across the diode and opamp together would be about 6pF so the increase in voltage should be measurable.
How did you work this out?The capacitance with the bullet in the center should be about 2pF - 10pF (larger for longer and wider bullets) and the capacitance across the diode and opamp together would be about 6pF so the increase in voltage should be measurable.
This might work well if typical CG projectiles had high permittivity but they don't. CG projectiles (for this kind of CG) are designed for high permeability - i.e. to be pulled by a magnetic field, not an electric field.
Besides, you've got a big magnetic pulse going. This is likely to swamp any very small signals.
It's a nice idea for being different but I think I'd stick with photogates.
Re: Coilgun projectile detection
Brendon, Mon Feb 12 2007, 11:30PM
But the projectile is iron so it conducts electricity and if PVC is used for the barrel it will have a dialectric constant of around 3 - 6. So when the projectile is inside the sensor it is actually two capacitors in series. I have no idea how much effect the magnetic field would have on it though.
Edit: Here's how I estimated the capacitance
C = e(A/d)
A = 2 * 0.008 * pi * 0.03 * 1/2 = 0.00075 (0.8cm diameter 3cm long projectile )
d = 0.001 (1mm walls)
e = 3 * 8.85e-12
C = 0.0000000000199125 = 20pF
Series capacitance is 10pF, I think in reality it will be less than this.
Brendon, Mon Feb 12 2007, 11:30PM
Simon wrote ...
This might work well if typical CG projectiles had high permittivity but they don't. CG projectiles (for this kind of CG) are designed for high permeability - i.e. to be pulled by a magnetic field, not an electric field.
This might work well if typical CG projectiles had high permittivity but they don't. CG projectiles (for this kind of CG) are designed for high permeability - i.e. to be pulled by a magnetic field, not an electric field.
But the projectile is iron so it conducts electricity and if PVC is used for the barrel it will have a dialectric constant of around 3 - 6. So when the projectile is inside the sensor it is actually two capacitors in series. I have no idea how much effect the magnetic field would have on it though.
Edit: Here's how I estimated the capacitance
C = e(A/d)
A = 2 * 0.008 * pi * 0.03 * 1/2 = 0.00075 (0.8cm diameter 3cm long projectile )
d = 0.001 (1mm walls)
e = 3 * 8.85e-12
C = 0.0000000000199125 = 20pF
Series capacitance is 10pF, I think in reality it will be less than this.
Re: Coilgun projectile detection
Simon, Wed Feb 14 2007, 02:03AM
That's a good point about conductivity and more interesting. Still, conductivity of projectiles is something to be avoided as eddy currents waste power and repel the projectile. A high resistance projectile will mean a large time constant in the whole RC system so you might not get good slew (reaction times).
I'd still stick with photogates.
The magnetic pulse will generate it's own currents in any metal around. Trying to measure a small voltage under these conditions will be messy.
Simon, Wed Feb 14 2007, 02:03AM
Brendon wrote ...
But the projectile is iron so it conducts electricity and if PVC is used for the barrel it will have a dialectric constant of around 3 - 6. So when the projectile is inside the sensor it is actually two capacitors in series. I have no idea how much effect the magnetic field would have on it though.
But the projectile is iron so it conducts electricity and if PVC is used for the barrel it will have a dialectric constant of around 3 - 6. So when the projectile is inside the sensor it is actually two capacitors in series. I have no idea how much effect the magnetic field would have on it though.
That's a good point about conductivity and more interesting. Still, conductivity of projectiles is something to be avoided as eddy currents waste power and repel the projectile. A high resistance projectile will mean a large time constant in the whole RC system so you might not get good slew (reaction times).
I'd still stick with photogates.
The magnetic pulse will generate it's own currents in any metal around. Trying to measure a small voltage under these conditions will be messy.
Re: Coilgun projectile detection
Dave Marshall, Wed Feb 14 2007, 03:15AM
This is very similar to one of the designs I tinkered with a couple years back while I was looking for a way to detect a coilgun projectile in an opaque barrel with no modification of the barrel itself.
Eventually it became pretty apparent that any design that was near the propulsion coil and relied on an increase in voltage to operate was impractical because of the induced currents produced by the firing of the propulsion coil(s) imitating the pvoltage created by the projectile. I also tried using the projectile as a moving transformer core, allowing it to induce current on a small coil wrapped around the barrel. This fell victim to the same problem.
The design I'm using now has been tested on a couple low power coilguns and so far has performed reliably. It can theoretically be done two ways, but I'll focus on the one I've found to be most practical so far.
Essentially it is a fixed oscillator running at 10Mhz (the higher the better to a point, I'll explain that in a minute). The oscillator signal is fed into an LC tank circuit made up of a small variable capacitor, and a coil of wire wrapped around the barrel. The coil and capacitor are tuned to resonate at the frequency of the oscillator so the 10Mhz signal can pass unattenuated through the tank. The output of the tank circuit is then fed through an ultrafast rectifier, a small smoothing cap (doesn't take much to smooth 10Mhz) and to a comparator. When the projectile enters the sensing coil, it detunes the tank circuit, causing significant attenuation of the 10Mhz signal, causing the opamp to see a drop in voltage.
With a variable reference voltage (and a stable power supply!) the comparator's trigger level can be set to a point that ensures mutual inductance with the propulsion coil doesn't cause enough attenuation to trigger it. This is pretty much impervious to induced current from the propulsion coil because even though there is plenty of voltage induced on the sense coil to fry the comparator, it isn't 10Mhz, and is substantially attenuated by the tank circuit. Since the comparator is looking for that decrease in voltage, it ignores any small spikes that pass through the tank circuit.
A higher frequency makes filtering any induced voltage spikes easier, but also makes the setup more prone to mutual inductance with the propulsion coil causing early triggering. Higher frequency also allows for a more compact tank circuit, so a balance has to be reached. When properly tuned, the sense coils can be wound directly UNDER the propulsion coils with the 500 joule per stage coilgun I tested the trigger with.
So there you have it. The Marshall coilgun trigger! (Seriously lame name, someone suggest something better...please!) Many thanks to Dave B and Chris R. who helped me through the long search for a usable trigger.
Dave
Dave Marshall, Wed Feb 14 2007, 03:15AM
This is very similar to one of the designs I tinkered with a couple years back while I was looking for a way to detect a coilgun projectile in an opaque barrel with no modification of the barrel itself.
Eventually it became pretty apparent that any design that was near the propulsion coil and relied on an increase in voltage to operate was impractical because of the induced currents produced by the firing of the propulsion coil(s) imitating the pvoltage created by the projectile. I also tried using the projectile as a moving transformer core, allowing it to induce current on a small coil wrapped around the barrel. This fell victim to the same problem.
The design I'm using now has been tested on a couple low power coilguns and so far has performed reliably. It can theoretically be done two ways, but I'll focus on the one I've found to be most practical so far.
Essentially it is a fixed oscillator running at 10Mhz (the higher the better to a point, I'll explain that in a minute). The oscillator signal is fed into an LC tank circuit made up of a small variable capacitor, and a coil of wire wrapped around the barrel. The coil and capacitor are tuned to resonate at the frequency of the oscillator so the 10Mhz signal can pass unattenuated through the tank. The output of the tank circuit is then fed through an ultrafast rectifier, a small smoothing cap (doesn't take much to smooth 10Mhz) and to a comparator. When the projectile enters the sensing coil, it detunes the tank circuit, causing significant attenuation of the 10Mhz signal, causing the opamp to see a drop in voltage.
With a variable reference voltage (and a stable power supply!) the comparator's trigger level can be set to a point that ensures mutual inductance with the propulsion coil doesn't cause enough attenuation to trigger it. This is pretty much impervious to induced current from the propulsion coil because even though there is plenty of voltage induced on the sense coil to fry the comparator, it isn't 10Mhz, and is substantially attenuated by the tank circuit. Since the comparator is looking for that decrease in voltage, it ignores any small spikes that pass through the tank circuit.
A higher frequency makes filtering any induced voltage spikes easier, but also makes the setup more prone to mutual inductance with the propulsion coil causing early triggering. Higher frequency also allows for a more compact tank circuit, so a balance has to be reached. When properly tuned, the sense coils can be wound directly UNDER the propulsion coils with the 500 joule per stage coilgun I tested the trigger with.
So there you have it. The Marshall coilgun trigger! (Seriously lame name, someone suggest something better...please!) Many thanks to Dave B and Chris R. who helped me through the long search for a usable trigger.
Dave
Re: Coilgun projectile detection
Brendon, Wed Feb 14 2007, 04:14AM
I'm going to go try it now. I haven't made a coil gun yet so I will try it on my potato gun.
Good point also. I'm not going to be using any complex projectiles to start with but a way of avoiding eddy currents and retaining conductivity could be to have the outside of the projectile conductive nearly all the way around. Leaving a gap running lengthwise. (Too complicated for me though)
Thats the main reason I want to try this. :)
Mod Edit: merged double post.
Looks like induced currents from the coils wont pose a significant problem. There are much bigger problems with this already. It's pretty obvious now that I have tried it. Moving my hand anywhere near the positive plate induces voltages and triggers the opamp. The positive pin on the opamp is basically floating in this design.
Time to go back to the drawing board.
Update : I've tried an Dave's method but using variable capacitance. The output from a 4Mhz oscillator is fed into a high pass filter. Without the projectile in the sensor the cutoff frequency higher than 4Mhz. It all works fine in SPICE but not in practice. I don't have an oscilliscope to check the the frequency is right and my multimeter doesn't measure inductance or capacitance.
schematic
Brendon, Wed Feb 14 2007, 04:14AM
I'm going to go try it now. I haven't made a coil gun yet so I will try it on my potato gun.
Simon wrote ...
That's a good point about conductivity and more interesting. Still, conductivity of projectiles is something to be avoided as eddy currents waste power and repel the projectile. A high resistance projectile will mean a large time constant in the whole RC system so you might not get good slew (reaction times).
That's a good point about conductivity and more interesting. Still, conductivity of projectiles is something to be avoided as eddy currents waste power and repel the projectile. A high resistance projectile will mean a large time constant in the whole RC system so you might not get good slew (reaction times).
Good point also. I'm not going to be using any complex projectiles to start with but a way of avoiding eddy currents and retaining conductivity could be to have the outside of the projectile conductive nearly all the way around. Leaving a gap running lengthwise. (Too complicated for me though)
Dave Marshall wrote ...
This is very similar to one of the designs I tinkered with a couple years back while I was looking for a way to detect a coilgun projectile in an opaque barrel with no modification of the barrel itself.
This is very similar to one of the designs I tinkered with a couple years back while I was looking for a way to detect a coilgun projectile in an opaque barrel with no modification of the barrel itself.
Thats the main reason I want to try this. :)
Mod Edit: merged double post.
Looks like induced currents from the coils wont pose a significant problem. There are much bigger problems with this already. It's pretty obvious now that I have tried it. Moving my hand anywhere near the positive plate induces voltages and triggers the opamp. The positive pin on the opamp is basically floating in this design.
Time to go back to the drawing board.
Update : I've tried an Dave's method but using variable capacitance. The output from a 4Mhz oscillator is fed into a high pass filter. Without the projectile in the sensor the cutoff frequency higher than 4Mhz. It all works fine in SPICE but not in practice. I don't have an oscilliscope to check the the frequency is right and my multimeter doesn't measure inductance or capacitance.
schematic
Print this page