Removing intact coils from very cheap speakers

S

starhawk

[H]F Junkie
Joined
Oct 4, 2004
Messages
8,908
I've got this neat idea for a project that needs a quartet of 8ohm coils within it -- not as part of speakers, though.

Since 8ohm coils on eBay are ridiculously large and expensive compared to what I'm looking for, I had the idea to buy this lot of four mylar speakers and get the coils out, or at least get the clear plastic bit (is it called the cone? I forget) off the front.

The coils obviously need to be completely intact...

Is that a fairly easy task that I've described, or am I asking for a painful pile of trouble?
 
As an eBay Associate, HardForum may earn from qualifying purchases.
So if each one is getting PWM off a 555 circuit, how long can I run them, roughly? Not really trying to rail them here, just generating enough EM to move a nearby wire.

I'm getting my inspiration from a recent Hackaday post :p
 
I assume its the piano one... If so, the hack was run with a 100w amplifier, so those would be nowhere near enough. I'd wind them myself IMHO...

The author wired 6 28ish ohm coils (cant remember) in parallel so they would all share the load placed on em... Those tiny speakers use every small gauge wire... Running 4 8 ohm coils in parallel to load share get you nowhere near the power handling, and peg you at an ohmage most amps can't support.,
 
Nope :p magnetophone. Not sure how I'll get the strings tho... I can't imagine they're cheap...
 
1 pound of 22 gauge magnet wire is about 500 feet long and has a resistance of 8 ohms. Just buy a 1 lb spool and you're done.
Or a half pound of :
26 gauge wire will be about 26 ohms
28 gauge wire will be about 65 ohms
30 gauge wire will be about 165 ohms
A 555 can probably only put out about a watt by itself. Any wire up to 30 gauge should be fine with that.

Edit:
Looking at the magnetophone project, and seeing you want to use a 555 for tone generation, I'd recommend approximately 60 ohm coils based on the 555 datasheet specs (15v V+, 200ma current). 400 feet of 32 gauge wire will be about 65 ohms and a half pound spool should have about 2500 feet.
 
Last edited:
I need something a little... smaller... than a pound of wire.

I think I'll just do the speaker-breaking thing. It doesn't seem too hard ;)
 
Assuming you use a 555 by itself, 8 ohm speakers will only give you about 1/6 of a watt. 60 ohm coils will result in about 1.25 watts. If you boost the current of the 555 (easiest way is with a transistor like a TIP3055) you can boost the power to an 8 ohm coil and get maybe 5 watts.
 
Those speakers I'm pulling apart are 0.25w speakers ;) are they not strong enough to affect a steel string in close proximity?

Basic idea, pulled from the magnetophone (as mentioned above) is to have four strings and four coils (as opposed to the Magnetophone's 12). A simple tunable 555 circuit (no need for fancy microcontroller crap IMO) outputs PWM to the coils, which vibrate the strings using EM, producing sound. The circuit I have now uses one coil (I could drive two coils per string if I had to, using a transistor -- how would you feel about using 2n2222's for this?) and two switches in parallel (one momentary and one latching) so that if one wants to play a note for a goodly while, one can do that without leaning on a button for too long.
 
2n2222 can sink about twice the current of a 555, so it's a small improvement. 0.25w speakers (assuming they are actually that rating) probably have about 42 gauge wire and won't like square waves. They might not last too long so watch for burning out.
Another idea might be to add a volume control to each one to give more flexibility in the sound, but that's something you do after you get it working
 
Oooh, volume control, why the fuck hadn't I thought of that?! Geez.

OK, if they don't like square waves, what shape waves do they like?

EDIT: also, if the issue is heat, is there a way (lol) to heatsink a coil...?
 
They like sine waves. Square waves will work, but will be less efficient and cause more heat build-up. The easy solution to both problems (waveform and heat) is to use a bigger, more robust coil.
 
Found a circuit I can use that will output sine waves :) it's more complex than I'd like but it will work.
 
Since he's gutting the speaker for the coil, all the physical properties of a speaker are no longer applicable. Only the resistance and inductance remain and in this application they can be considered constants, so no impedance.

starhawk - you don't NEED sine waves and even with perfect sine waves you can still easily blow the speaker coils. If you just don't want to buy new wire, look for old motors, alternators, transformers, etc that are being thrown away for sources of magnet wire.
 
Fenris_Ulf said:
Since he's gutting the speaker for the coil, all the physical properties of a speaker are no longer applicable. Only the resistance and inductance remain and in this application they can be considered constants, so no impedance.
Click to expand...

Yep, but the link provided is a good start in helping to understand this just in case it's needed.
 
@Fenris_Ulf -- I don't fancy spending a few hours/days/weeks/whatever winding a few hundred turns of wire in a circular motion, is the problem. I don't care where the wire comes from (right now at least) -- I just don't want to be the one assembling the coils.

What about the guts of a 2" PC speaker? I've got two or three of those, if they're not hard to gut...
 
You do realize that I can get speakers cheap, right? I have more than enough time to get them into pieces...
 
I just throw ideas on the off chance they might be helpful. It's your project and I wouldn't presume to know what you can and can't do or what might or might not work. I will be looking out for you to post an update if you come up with something really cool.
 
The other bit of info -- look at the price for the speakers I was going to get. Link is in the first post.

Now look here, for the cost of a 2" PC speaker --> http://www.ebay.com/itm/300813015580

Voice coils? Much more expensive... and too big physically as well...
1 --> http://www.ebay.com/itm/271001339021
2 --> http://www.ebay.com/itm/280906150762
3 --> http://www.ebay.com/itm/130679802918
4 --> http://www.ebay.com/itm/130766840796

I hardly ever have much money... but I have time in bulk freight quantities. Hence why I'm more interested in chopping up speakers than in buying more-expensive coils and adjusting my design to compensate for the coil size increase.
 
As an eBay Associate, HardForum may earn from qualifying purchases.
@Fenris_Ulf -- hence why I'm leaning toward the mylars, or rounding up a few extra 2" PC speakers and trying to not wreck the coils while disemboweling the everything else ;) BTW, four of those transducers, plus cheapest available shipping, is just over $20, which is more than I want to spend by at least double. I appreciate the effort, don't get me wrong, but that's too much money right now.
 
If you really have more time than money, I'd lean toward salvaging some wire from transformers, magnetrons, motors, alternators, etc. If you've got a decent kitchen scale, you can just wind by weight rather than measure length to get the right resistance.
 
could also possibly check out clearance transformers.. they can be used as inductors too. Parts Express has had some reasonably powerful ones for cheap on clearance in the past - not sure about now.
 
Picking this back up. I have four coils from... oddly enough, a motor from a PC drive of some sort. I don't know what drive, anymore, since I pulled them at least a decade ago if not more.

My SparkFun Digital Multimeter says that, depending on how steady my hands are and how I hold the probes, the coils have ~3.5ohm - 5ohm impedance/resistance. It seems to like 3.7ohms, so I assume that they're basically 4ohm coils. Copper wire, thin but nowhere near voice-coil thin (that shit is tiny! I took apart a PC speaker; it didn't go as well as I'd hoped...). No idea what AWG.

I have a number of questions but the most important two are --
(1) how will a ~4ohm coil behave differently from a ~8ohm coil?
(2) what's the best way to drive these, assuming that I'm sending them PWM from an Arduino Uno chip (ATMega328)?

EDIT: a friend of mine tells me that these coils are HDD-head servo coils. They most likely came out of an MFM hard drive that I took apart fifteen or so years ago, not yet knowing what such a thing was. LOL.
 
Drive them with a transistor or MOSFET. What voltage are you driving from? That will determine how much current you need.
 
Standard 5vTTL is what I'm planning on.

I'm thinking about using a 555 circuit instead tho. Not sure what voltage is best for those -- I don't use 'em too often.
 
555's are usually rated up to 12V, though I don't know offhand what kind of output current they can source/sink.
 
555s are good for 200ma either sourced or sinked. With a 12v supply a 60 ohm coil would be optimum (post #8 and #10). With a 4 ohm coil you should get maybe 0.4 watts. That might be enough. If not, a transistor or mosfet could be used to boost the current.
 
I'll probably be using a transistor of one sort or another then. I'm also exploring PICAXE chips (which I have a pile of) to see if I can use one for this.

This is getting interesting.
 
You'll probably need about 2 or 3 amps from the transistor. If you figure your microcontroller can sink 2ma, you'll need a Vbe (current gain) of 1000. For this I'd recommend a Darlington (a cascaded) transistor. A 2n2222 and TIP3055 in a Darlington arrangement would work as well. A 555 could just use a power transistor like the TIP3055.
 
I think I'm leaning towards 555's. The only reason for microcontrollers here is some secondary stuff that I don't think I'm smart enough to pull.

Basically what I'm picturing is a sequencer type thing -- I want to be able to record and loop button sequences in real time (so record some button-presses as they happen, but also make the associated PWM outputs /at the same time/ -- and then, on command, play them back in a loop until told to stop). But that's probably really complicated and I'm not experienced enough with PICAXE stuff to pull that off. Arduino will be a doozy at best -- I seem to have a system-level incompatibility with C and any C-derived language.
 
Maybe try some kind of I/O using USB - I've used the USB Micro u421 with good success.
 
That's WAY over my head.

My current experience with microcontrollers is limited to the PICAXE itself -- which is programmed in BASIC, FFS! -- and the one project I've done is a simple program that outputs ASCII over TTL to a compatible LCD. Sort of a nerdy name badge thingy.

Yeah, 555's are sounding better every second. It's more limited -- but I can always make it better later on down the road...
 
You could make it with 555s and later on add an interface with the u421. The input lines could read button presses and the output lines could drive relays that push buttons. Easy enough to record button states and play it back that way.
 
You must log in or register to reply here.
Back
Top