Page 1 of 2

How to best control and electromagnet with sound

Posted: Sat Feb 17, 2007 3:27 pm
by rkennedy
O.k. I'm new to this, but I think I am starting out slowly enough with this one...

I want to do a project where sound controls and electromagnet, which in turn controls some magnetic fluid. Nothing fancy here. Just Audio ---> Magnet.

Here is what I think I know so far.... :)

A voice coil for a speaker is just a basic electromagnet. Via some testing, it doesn't seem to be a very powerful or efficient one. It doesn't need to be since it just bounces off the string permanent magnet in the speaker housing.

I have also deduced that a moderate audio amp will produce up to 6V A/C on the outputs. This of course varies with the input.

I have also found for sale some pretty powerful (they claim a 200lb holding force) electromagnets that run off of a single D cell. Which is IIRC 1.5V max.

Therefore I should be able to power an electromagnet off of an amp and get what I want....


I am concerned that my HZ are way to high to effectively move the ferrofluid. All tweeter and no bass if you get my drift. It seems like I would need to bring the Hz way down to push around this stuff.

Any help here?

Posted: Sat Feb 17, 2007 3:52 pm
by dacflyer
whats the purpose ? what are you trying to build ? whats it supposed to do ?

i have seen some metallic fluids , basically they are just like steel powder in a oil like mixture..

tell us more info on your desires what your trying to acomplish..

Posted: Sat Feb 17, 2007 4:02 pm
by Chris Smith
There are no free lunches in physics.

A large force from a small voltage MUST lose in other ways, and that said,... response time would be the first loss encountered.

Six of one, half dozen of another, keep in mind that any gain must accompany some kind of loss.

Posted: Sat Feb 17, 2007 4:02 pm
by L. Daniel Rosa
I think he wants noise to summon the shiny porcupine demon from a pool of metallic oil, and silence to dismiss it.

Posted: Sat Feb 17, 2007 4:11 pm
by rkennedy
Basically that is it.

Noise = magnet on at full power

Some noise = magnet on at less than full power

Silence = magnet off

It seems like I may have to go DC with this though? Wouldn't A/C have too much "hum" to let the field completely collapse?

Posted: Sat Feb 17, 2007 5:55 pm
by Robert Reed
As Dac said, we need more info, but if I understand you correctly, here are my thoughts:
You will never get the power out of a standard Op-Amp to control an electromagnet.
You will probably have to convert audio AC to DC, this will slow the response down considerably, but is simple to accomplish. If all you want is to increase magnet power vs. audio power, this would work fine as it is fast enough to track varying audio levels.
If you want to track audio frequencies,this will not work fine, as its response is too slow to keep up with that.
If you want to track bands of audio frequencies in the audio spectrum.this will work for that, but you will need various IC filters in the circuit.
If you decide to go the DC route,then obtaining adequate power to drive the magnet is easy.
And finally,until we know exactly what your needs are,its just a guessing game at this point.

Posted: Sun Feb 18, 2007 1:27 am
by jollyrgr
I think what you want is called a SINGLE CHANNEL COLOR ORGAN. Color organs are "light show" devices that turn on colored lights. A three channel color organ ( what you will see the most of) will have different colored lights that repsond to different frequencies. You mentioned that [sic] "I am concerned that my HZ are way to high to effectively move the ferrofluid. " I'm assuming audio frequencies are involved here and this is the only reason I mention the three channel color organs. The basic action is the louder the noise the brighter the light. Many AC power color organs (at least the schematics I've seen) use SCRs in their output; perfect for a magnet (unless you are trying to go with a motor). So be careful if you jump into one of those schematics.

What I think you are asking for is a simple variable VOX circuit. VOX stands for Voice Operated Switch. These are super simple to make from a couple transistors. If you are only have about DC output of 1.5V max, even easier. What I'd do is get something like a LM386 audio amp chip and build the basic audio circuit (if you want to go with a microphone and/or an isolated amp). Take the output of the amp and rectify it with a common diode such as 1N4002. Take this output and filter with a small electrolytic capacitor, such as 250uF from output to ground. Feed this DC output to a potentiometer's end connection. Connect the wiper to the BASE of a power transistor such as a 2N3055. Connect the COLLECTOR of the 2N3055 to the Vcc. Connect the EMITTER of the 2N3055 to the "input" of the electromagnet. Connect the other wire of the electromagnet to ground. See this schematic:


This is a "general" circuit as you have not provided enough detail as to what you really want to do. This circuit is untested but should work. Any comments, suggestions, or corrections are welcome.

Posted: Sun Feb 18, 2007 9:44 am
by rkennedy
Like I said, I am a bit new to this, so pardon any communication problems.

Here is exactly what I want.

I want to take audio input from a standard MP3 player via the headphone jack and use that to create a varying magnetic field which will cause a pool of magnetic fluid to respond in time with the audio.

This is my first real attempt at this, so even if I can just get the music to pulse in response to volume (when the audio is loudest, the magnet is strongest) or to frequency (lows do one thing, highs do another).

I am looking for baby steps here. Nothing too fancy to start with. Simple is the key. Once I have that down, I can get fancier I figure.

Ideally, the color organ sounds the way to go, but here are the concerns I have been told to look out for in any design I go with -

1 - The magnetic fluid has an inherent mechanical limit as to how fast it can respond to a signal. This could be a problem if my signal frequency is too high, the fluid itself will not be able to respond. Imagine slowly moving your hand up and down versus rapidly moving it up and down. If you move it too fast you won't even see your hand move! I am going to see if I cant run some tests tonight on an AC electromagnet I built last night. It has that inherent 60hz "hum" to it. Maybe that will give me a better idea of the mechanics I am dealing with.

2 - I need enough of a magnetic field to saturate the magnetic fluid and get it to respond. I am assuming that a well-made 1.5v electromagnet that I have seen advertised in a few places will do the job. It claims to hold over 200lb (I figure less in real life).

Now, the color organ sounds like a real possibility, but I am a bit wary of the output. Many kits seem to have a simple 120v output. All I know of 120 and electromagnets is that if you plug one into your 120 wall socket, it will make a nice batch of toxic smoke for you, and maybe a small fire. Would this be an issue in the case of a regulated system?

Thanks again for all the info. You are actually being quite helpful.

Posted: Sun Feb 18, 2007 2:31 pm
by dacflyer
Hmmm ok, then try this...

the audio is basically a/c in a sort of way...and you want it to become DC?

try a full wave bridge rectifier..a/c goes in....dc comes out..
try that...
let us know the outcome good luck

Posted: Sun Feb 18, 2007 3:52 pm
by rkennedy
I'm going to get a rectifier this week, but can DC do the job? I was lead to believe that only AC can deal with audio signals.

Posted: Sun Feb 18, 2007 5:07 pm
by dacflyer
well audio ac would cause the electromaganet to work both ways againts a maganetic object such as a maganet...
but in dc the coil would only work in one way and vary with voltage..
also the effect would be seen with the polarity of the connections..

are you wanting to put like a coil of wire in a magenetic fluid and see the effects or ??
i am trying to picture what your gonna make :)

Posted: Sun Feb 18, 2007 6:43 pm
by Robert Reed
Quote " I 'm going to get a rectifier this week, but can DC do the job? I was lead to believe that only AC can deal with audio signals."
For starters, The audio level you are dealing with will be too low for normal rectifier operation, let alone a bridge rectifier. You will need amplification to get that signal up to several volts before you do any processing. Diodes don't start to conduct until the voltage across them reaches about 0.7 volts, and any signal of less than 5 volts p-p would not give very good linearity. There is an op-amp circuit known as "the ideal rectifier" and also known as an absolute value circuit. These will give almost perfect lineaity from a few millivolts on up. You can also incorporate gain into these circuits or add a stage after. I have never worked with ferro magnetic fluids, so I have no idea what the parameters required would be. I think a good place to start your project would be to "play" around with the fluid with a crude electromagnet. You can manually switch the magnet on and off at different rates and note the effect. Again, if you want the fluid to track the program level, this should be doable as the response time would probably be in the high fractions of, or low seconds of the program level. If you want to split things up in regards to frequency (i.e.- bass drums or brass) you will need some op-amp pass filters (these are the easiest) in the circuits front end. The easiest part of all is to get the required power after all this processing is done. As far as keeping step with actual ac frequency, I doubt if the end components (electromagnet,fluid) would even come close to responding. As far as keeping it simple, it never seems to happen in the real electronic world :sad:

Posted: Sun Feb 18, 2007 10:21 pm
by jollyrgr
Please review my schematic above. This is about as simple as it gets. I understand you wanted to rid yourself of the high frequency component and make it simply a sound level device. What frequency are you sending these signals? If it is an MP3 player the highest sample rate most can take is 44kHz, if that. This means your highest audio frequency will be 22kHz. If so you may need to take a different amp then the one I show. But if this is something you can hear, the circuit should work great. I have used the simple diode and filter capacitor trick numerous times for a VOX circuit. In fact I used a similar circuit (but an LM324 OP amp) to drive a light in a pumpkin. (The audio source was a sound chip.)

In my schematic take a look at the diode; this converts the AC signal to DC. An MP3 player may work on 1.5V to 3V DC. The output to the headphone jack will not be more than that. Usually the current level is quite low. Even assuming you do get 3VDC out of the headphone jack this would be into a 32 Ohm load. You will not even get 100mA out of the device! You need an amplifier. The LM386 is acting as a preamplifier to the 2N3055. A 2N3055 can handle several amps (more than a D battery will even supply).

The 200LB electromagnet you speak of may very well hold 200lbs. But there is a "trick" to it. The magnet only will hold 200lbs when mating to a machined surface. Usually these come in two parts. The magnet part with an eye bolt and a flat metal surface with an eye bolt. Put in your D battery and place the two surfaces together. Likely you won't pull them apart linearly until you disconnect the battery. They use similar things to hold doors shut in security situations. These magnets run on 24VDC and when the two surfaces are together you cannot pull them apart. A door will remain locked shut. But only when the mating plates line up. You can put a screwdriver or other metal object on the magnet and easily pull it away.

Posted: Mon Feb 19, 2007 7:30 pm
by Robert Reed
Quote-The 200LB electromagnet you speak of may very well hold 200lbs. But there is a "trick" to it. The magnet only will hold 200lbs when mating to a machined surface.

How very true and how very expensive this is-'mating surfaces'.I once saw a demonstration of two non magnetic but close grained metals whose surfaces were perfectly machined (micro inches) that once pressed together no reasonable force could get them apart. Something to do with surface tension,

Posted: Mon Feb 19, 2007 8:19 pm
by Externet
I believe the 'electromagnet' word is flawed to achieve the effect.
A viscous fluid makes it worse.
A voice coil alone, no core, with the fluid as core should give a possibility to evaluate the ferrofluid response to frequencies and amplitude, trying lighter viscosities.

Ferroparticles in alcohol could yield nice fast response, and it is transparent, to visualize better.
As the ferro particles would not stay suspended in a non-viscous fluid, I would try the coil wound externally to a clear shallow round container, to be visualized from above.
Again, no iron nor magnet core in my opinion.