1
\$\begingroup\$

I am designing something similar to a coil gun. So what I have is a magnet passing through/adjacent to a coil. Due to the electromagnetic induction, I do get a pulse(triangular) across the coil. However, the amplitude of this pulse is very low to detect, as its around 60mV peak.

I tried using a comparator(LM311) for detecting that low voltage, but noise disrupted the operation, as the noise level was also high. Is there something I can do to detect that pulse?

Things I thought about : 1. Increasing windings on the coil 2. Using a stronger magnet but apart from these, is there something that can help me achieve the same?

Currently the coil used is a small sized with about 30 turns of around 23-25 gauge magnet wire

UPDATE: I tried increasing the resistance by adding a 1Kohm-100Kohm in series to the coil, but it did not increase the voltage much. I also tried winding a coil with more turns but still the max voltage was around 40mv. Also, here is a video: http://www.youtube.com/watch?v=hajIIGHPeuU I came across. The current generated in this case is 20uA, which for a resistance of 1Kohm, should give a Voltage of 0.02V. Any other way, I can increase the voltage level. Would having a core help in anyway?

\$\endgroup\$
  • \$\begingroup\$ LM311 does not seems a low noise, fast op-amp for the application. Also the noise is only at the discharge, or is continuous? What is power supply? \$\endgroup\$ – Diego C Nascimento Jan 10 '14 at 22:26
  • \$\begingroup\$ 60mV isn't that low so I'd want to know where the noise is coming from - maybe you can supply an oscilloscope screen shot or a circuit of detector. BTW, to measure a pulse amplitude like this a comparator won't really be any good. A picture of the pulse would be useful for helping you. \$\endgroup\$ – Andy aka Jan 10 '14 at 22:57
  • \$\begingroup\$ I will post a screenshot tomorrow. \$\endgroup\$ – Sherby Jan 10 '14 at 23:32
  • \$\begingroup\$ You say passing though or adjacent to a coil. There is quite a difference. Can you describe the geometry? As you say the main things are the number of turns and what Faraday would have called the rate at which "field lines" are being cut by the wire. Use magnet wire like #38 with a lots of turns and maximize the change in magnetic field through the geometry of the setup. \$\endgroup\$ – C. Towne Springer Jan 11 '14 at 5:38
  • \$\begingroup\$ For the geometry, initially I was planning to pass the magnet through the coil, but due to space constraints I ended up placing the coil to one side, so instead of through it, the magnet is now passing adjacent to it. \$\endgroup\$ – Sherby Jan 11 '14 at 6:19
0
+50
\$\begingroup\$

There is a little amplifier I am very fond of which I have used with dynamic mics. Since your coil situation is much like the one faced by someone trying to amplify a dynamic mic, I encourage you to try this circuit and see if produces a usable signal. Just replace the "microphone" (speaker) with your coil.

http://electrosuite.com/audio/bc547-dynamic-microphone-amplifier.html

The lower transistor acts as a common base amplifier, which is fairly unusual. It has a current gain of less than 1, but it does provide significant voltage gain. The upper transistor is in a common collector configuration, so it has essentially unity voltage gain, but the current gain is good. In short, this first amplifies the voltage, then the current, to give you a strong output signal. Although you will likely need a second stage after this, that should be much easier.

EDIT:

Concerning your noise issue, do you know where the noise is coming from? Sometimes I just have a power supply that's super noisy and ruining everything. I like low budget solutions, because of, well, my budget, so I usually test this by running the circuit on batteries. If things start magically working, then it's time to work on the power supply.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ It did amplify the pulse, but added a very noisy DC offset to it. The noise is a sine wave of high frequency(ringing), of delta 350mv. However the pulse is amplified with a peak(from the delta) of 300mv, which is good enough. Any suggestions, on how I can remove the DC offset? I tried adding a coupling cap, but it would block the DC. \$\endgroup\$ – Sherby Jan 16 '14 at 2:13
  • \$\begingroup\$ Hmm... I have often found, with making amplifiers, that when you get inductors and capacitors together in a circuit, they will usually find a way to oscillate despite your best efforts. I have found that "when in doubt - add capacitors" is a good policy. Perhaps a small capacitor between one of the transistor's bases and ground will fix that noise. As for the DC offset, I am somewhat confused. You seem to find blocking the DC offset desirable ("how I can remove the DC offset"), but then you say, about adding a coupling capacitor, "but it would block the DC". I'm not sure what you mean. \$\endgroup\$ – Void Star Jan 16 '14 at 2:32
  • \$\begingroup\$ Turns out removing the capacitor between the output and the resistors(top one), actually fixed the ringing. I was planning to use the coupling capacitor to give a negative/downward offset. \$\endgroup\$ – Sherby Jan 16 '14 at 3:46
  • \$\begingroup\$ Well, removing that cap is a neat trick. I'll remember that if my amp ever turns into an oscillator. I don't have the right parts (remember the budget issue?) so I use 47uF caps in place of the 22uF caps. You could try that for kicks, but since it's working now there's not much point. I think the purpose of that capacitor is to improve linearity. \$\endgroup\$ – Void Star Jan 16 '14 at 4:32
  • \$\begingroup\$ Actually, I change my mind. Won't it give the circuit some positive feedback, increasing gain? \$\endgroup\$ – Void Star Jan 16 '14 at 4:33
2
\$\begingroup\$

I think amplifying the 60mV pulse right away from the coil output by a good and well designed amplifier which also should be low noise one , this will help the 60mV pulse signal to be strongly above any kind of noise and so will be easily detected.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ What can I use for amplifying the small voltage? (apart from Boost converters). \$\endgroup\$ – Sherby Jan 14 '14 at 1:48
  • \$\begingroup\$ I wouldn't call it a boost converter -- those are designed to supply power, not condition signals. The common word you are looking for is probably 'amplifier'. The EE part could be a transistor or an op-amp most likely. \$\endgroup\$ – HL-SDK Jan 15 '14 at 16:43
1
\$\begingroup\$

A current is induced in a wire moving relative to a magnetic field. At the moment you're measuring the voltage generated by that current across the impedance of your coil. If you add a resistor in series with the coil, and measure the voltage differential across the resistor using an instrumentation amp (which is not a regular op-amp or comparator), you may find both noise rejection and signal intelligibility improve considerably.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ I'm pretty sure Faraday's Law tells you the voltage produced when X amount of magnetic field changes with respect to Y coil winding. To get more V will take more magnetic change/time or more coils. see hyperphysics.phy-astr.gsu.edu/hbase/electric/farlaw.html \$\endgroup\$ – George White Jan 11 '14 at 1:12
  • \$\begingroup\$ That makes a lot of sense. I was aware of current being produced, but did not think of it as this way. I will give it a try tomorrow, and will let you guys know. \$\endgroup\$ – Sherby Jan 11 '14 at 2:33
  • \$\begingroup\$ Increasing the resistance did not help. I hav updated the question. \$\endgroup\$ – Sherby Jan 13 '14 at 18:32
0
\$\begingroup\$

The antediluvian and inexpensive LM311 is perhaps not ideal for this application, but it should work, given a reasonable circuit and pulses that are in the microsecond range or longer.

First, make sure that one end of the coil is solidly grounded so that any common-mode noise that may be capacitively coupled is conducted directly to ground. Try connecting it to the inverting input directly.

Try a reference of about 30mV with perhaps 5mV of hysteresis (feedback from output to the non inverting input and a divider from the supply, so three resistors- for example 100 ohms to ground, 16K to +5 and 100K to the output (output high has a 4K7 pullup, say, to +5), assuming +/-5V supplies. Ground the output common. You can test this with a function generator (say a 1kHz 60mV sine wave). It should give clean output pulses.

Use positive and negative supplies (say +/-5V to +/-15V) to ensure that you're inside the common-mode range of the LM311 (or it won't work properly at all). It can be adapted for a single supply later, but first get it working.

If that does not give a clean pulse out, post an oscilloscope shot of the voltage on the coil with one end grounded to your circuit. If the pulse is not fairly clean at that point, then you've got a different problem.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Antediluvian... I had completely forgotten about that word. Thanks :D \$\endgroup\$ – Void Star Jan 16 '14 at 4:35
0
\$\begingroup\$

A coil gun isn't much more than a special case of a linear motor, which is just a special case of a brushless DC motor. In high performance motor applications, or ones where you must know where the magnetic field is at in order to drive it most efficiently, hall effect detectors are used.

I suggest that while you can use the coils for both feedback and driving, you might save time and frustration by simply adding hall effect sensors to the assembly.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Thanks Adam, I have considered that option. Since it is passing adjacent to the coil, the sensor is activated in both conditions of approach and leaving of the magnet. An actuation of the coil can cause attraction and may slow/stop the magnet. Using a coil, gives a positive and a negative pulse, signifying the approach and leaving of it. \$\endgroup\$ – Sherby Jan 15 '14 at 22:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.