Can Audio Waves Really Control Electromagnets for a Ferrofluid Sculpture?

[Rafflez]

Hi everyone,

I am a member of an educational group at my school and I proposed to my boss that we make a ferrofluid sculpture for future outreach events. She thought it was a great idea but now I have to figure out how to make it and I have no electrical experience. I would like to take an audio file (from an iPod or computer, maybe even a microphone) and turn that into variable voltage (or current) to make an electromagnet vary in strength. Can anyone help me figure this out?

 

[dlgoff]

If you want to make a really cool one, you should have multiple channels with a electomagnet for each. e.g. 3 audio channels; one for low frequencies, one for mid-range frequencies, and one for high frequencies. For this example, you would need to make a circuit with 3 band-pass filters with their outputs driving transistors capable of delivering enough current to produce a good magnetic field.

Your circuit would be similar to those used for http://en.wikipedia.org/wiki/Light_organ" .

Here's a 5 channel schematic showing operational amplifiers being used to make the band-pass filters and also includes the power supply and audio input circuitry.

http://www.xkitz.com/catalog/documentation/Light_Organ_Schematic_5_Channel_XLO-5.pdf"

 

[Rafflez]

That's exactly what I had in mind I just had no idea how to build it. What kind of hardware would I need to build that set up, and would I need some sort of converter or software so that the input is an MP3 or some other file and the output is just DC with the strength varying depending on the volume of each frequency range?

 

[dlgoff]

You could perhaps buy an audio equalizer and modify it to drive your magnets. I'm a believer in NOT reinventing the wheel. Keep it simple until you gain more electrical experience. And you should consider safety in your project. You don't want to get yourself or anyone else hurt. Unless you have experience with mains power, I'd say find some UL approved power source (a wall wort perhaps) to drive your magnets.

 

[Bob S]

I suspect that you want the ferrofluid to respond to the audio signal envelope (i.e., rectified audio signal), and not the audio waveform itself. There are two reasons for this:

1) Electromagnets generally have very poor frequency response, unless the core is laminated (like an audio transformer), or the core is made of ferrite, and

2) The ferrofluid is very viscous and may not respond to audio frequencies, unless the audio signal is rectified.

There are precision rectifier op-amp circuits that can be used to produce a good envelope signal. You will need many amp turns to do this, so a good audio amplifier (like one that can drive many watts into a 4 or 8 ohm speaker) may be required. Do some tests with an electromagnet and determine how many amps you need.

This website shows a simple and a precision audio VU meter circuit using diode rectifiers: http://sound.westhost.com/project55.htm

Bob S

 

[Rafflez]

Yeah that sounds a lot like what I would like to build. Sorry for sounding so uninformed: a rectifier will turn the audio signal from AC to DC (which is probably something I want to prevent the ferrofluid from splashing or being repelled, right?), and I want the ferrofluid to respond to signal envelope so that the ferrofluid motion is smooth and gradual.

If I were to build this using only one electromagnet -if it works well I will work my way up to more- what would I need?

1) I was thinking I could make my own electromagnet. Is that an ok idea if it's housed properly (this would probably be around kids)?
2) If I play music from my ipod would I just need to connect it to a "precision rectifier op-amp circuit" and the output control the electromagnet?

 

[Bob S]

You should be able to use the precision rectifier op-amp circuit (VU circuit Fig #2) in

http://sound.westhost.com/project55.htm

You will have to adjust the gain of the op-amps (VR1) to get proper voltage gain. You may need even more gain, depending on the maximum voltage output of your ipod (what is it?).

To get adequate current, you will need to replace the output transistor Q1 with a darlington (TIP120 or equiv) available from Radioshack. See

http://www.datasheetcatalog.org/datasheet/stmicroelectronics/4128.pdf

You should redesign the output circuit (TIP120 emitter circuit) for the magnet coil. For starters, the magnet coil should be about 30 ohms (1 amp at 30 volts, or 30 watts). Use 20 Ga. solid copper enamel coated "magnet" wire, (10 ohms per 1000 ft, rated at 1.4 amps). The +/- 15 V power supplies should be able to source / sink 1 amp. The TIP120 heatsink should be suitable for 15 watts minimum. You might also consider an open collector (common emitter) output circuit, which has more voltage gain.

Bob S

 

[Rafflez]

Sweet Deal! Thanks for the input, at least now I have an idea of what I need to do. I'm still confused by the schematics, but since this will probably be a summer long project I have plenty of time. I'll try to get in touch with some people at school to see if they can help me out, and if I have any questions I know where to go.

Rafal

 

[schip666!]

Definitely go with the Envelope Follower or Precision Rectifier idea -- you didn't mention multiple channels so all the talk of color organs and equalizers is for naught. You may not even need the 'precision' part of the rectifier if your signal is of a high enough voltage that the (approximately) .7v drop across regular diodes is 'down in the noise'.

The main issue is driving the electro-magnets. I've never worked with ferro-fluids, but in discussions with those who have I saw HUGE current requirements. Their driving coils were basically large insulated rolls of copper 'strapping', which means the resistance is quite low. I would wind up a candidate magnet and run it with, maybe, a car battery, just to see if I was in the ball-park for size and power. Once you have a working magnet system, you can figure out the power requirements of the transistor drivers.

I tried to find some refs for you but the best I could do on short notice is this:
http://pages.csam.montclair.edu/~ye...SilberKnobloch_FerrofluidPatterns_PhyaD88.pdf
where the interesting technical bit might be in an (apparently unpublished) paper by Crutchfield, Jeffries, Lam (ref #5)... Some digging on google scholar or other sites may turn up more.

 

[dlgoff]

...you didn't mention multiple channels so all the talk of color organs and equalizers is for naught.

Correct. But what kind of patterns are you going to get? Not to pretty.

One could certainly take the outputs of band pass filters and rectify to get an envelope signal. But I guess that would be too much to expect.

 

[schip666!]

But what kind of patterns are you going to get? Not to pretty.

My recollection is that one sees "magnetic lines of force" just like with iron filings. Most likely the amplitude of a typical musical signal would vary slow enough to make a nice pattern. Think following the beat of a kick drum, etc...

 

[dlgoff]

I'm thinking in terms of an orchestra.

 

[sophiecentaur]

The snag here is that you are trying to develop what could amost be called a 'new technology'. The problem is not in processing your sound signals to produce 'syllabic' information about the music content. The problem is in producing a significant visible effect this way.
I think you will be hard pressed to produce an effect this way which is anything like as impressive / pretty as the 'visualisations' you get on computer screens from iplayer and the like.
You will need a lot of power to drive any set of electromagnets hard enough to obtain a large area / volume of appropriate field strength.
The impedance of a magnet with enough turns of wire and a suitable core will be enormous. How would you intend to force a rapidly current of several Amps through it? You are talking in terms of a very high power amplifier.
I guess a matrix of many small electromagnets behind a shallow 'bath' of the fluid could give some detectable patterns and effects in two dimensions.