Increasing the Voltage Output from a Computer Speaker Amp

this all works fine, but i don't think the amp is providing 12VDC

That's correct. A standard computer amplifier (say 5W per channel) will put out around V=square_root(5W*8ohm), giving about 6.3V. Note that this is an audio signal and so the current (and audio signal) is AC, your 12VDC fans are designed for DC (like the output of a battery). If you add a full wave rectifier (bridge rectifier + capacitor) between the fan and the computer amplifier you will convert the AC signal to a DC signal and get around 7.4VDC peak, although if you make the capacitor too large it will average out all the AC peaks that make up the AC sound (and your fans move).

To keep it really simple, you cannot modify the computer's speaker amplifiers as they are not designed for that purpose, and changing the capacitors to larger ones won't give you more sound volume (and hence voltage for powering your fans). If you cannot turn your speaker volume up any higher (either with the physical volume knob or using the computers sound volume settings thru the desktop) then what you need is a bigger amplifier that you can connect between your computer and your fans. Say 15W per channel (fan) should be enough to give you up to around 12VDC during the sound peaks to drive the fan. The simplest option for you is likely to buy a set of computer speakers with larger amplifiers built in (before buying check the box they come in for details on the actual amplifier power).

Please note that your experiments may damage your amplifier and the fans (and possibly even your computer sound card), as they are not designed to be used this way. Just something to be aware of.

yeah, i have a bridge rectifier in place, had solved this problem a while back.

ok, well thanks. i have put in a lot of work onto this board and was wishing this would not be the only path possible.

Since i need a new amp i might as well make one and learn about this stuff instead of buying one. Anyone have schematics for a 15w amp?

EDIT...

would somthing like this work?

http://www.amazon.com/Pyle-PCA1-2x15W-Stereo-Amplifier/dp/B0012KZNP4

That looks suitable. You used to be able to buy electronic kit sets from stores that you could build yourself. I don't know if you have any stores nearby that stock kits but you could look on the internet for an amplifier kit set if you wanted a little more practical electronic experience (they are not too difficult).

Try an internet search of "amplifier kit" or similar. Someone else here may know of stores in the (USA?) that you could try.

You can't modify the audio amp specifically but you could break into the circuit somewhere and build a secondary power amp that would deliver close to 12v. You could probably hang an op amp circuit and a power transistor on the output of the speaker amp and drive the fans with that. There are a lot of circuits on the web that will do this, the trick is to have a power supply with a rail of at least 15Vdc and go from there. That will deliver 12v without clipping and the quality of the frequency spectrum won't matter to the fan. What else do you need?

as i said i did not want to buy one so excuse the edit i made.

i did search for 15w stereo speaker amps and found quite a few documentations on this, but i don't know which would be better for my application, so if some one who knows more about all this can show me some information that would be great, thanks

ps sorry for the double post

You don't need a great deal of amplifier options for your application. The key parts are that it has RCA plugs (or something similar) to easily allow you to connect to your computer's audio output built into your motherboard or separate sound card (check that your computer has RCA sockets for this purpose). You may need a sound splitter adaptor plug and socket (one RCA male plug to 2 RCA female plugs) if your computer only has a single mono RCA output socket (female socket).

Also, I would advise an amplifier that has an external plug pack (or wall wart) power supply. This will keep the dangerous mains voltage (110V, 220V-240V, etc depending on your country) out of harms way if you want to open up the amplifier for a look.

There are other options, such as connecting a home theater or midi system between the computer and the fans, but the danger and cost of damaging them is rather high (so I wouldn't recommend it).

http://store.qkits.com/moreinfo.cfm/FK607

would 2 of those get the job done?

edit...

or one of these

http://store.qkits.com/moreinfo.cfm/FK608

Yes, those are exactly what I was thinking of.

Yeah, that will do it.

This looks pretty good but wouldn't you get a switch on/switch off effect from the Op Amp dependant on the level of amplitude of the sound wave? Ideally what he needs is variable voltage dependant upon the amplitude of the sound wave and not drawing too much current from the amplifier as to blow it. I am not sure how he is ever going to get the fans to respond to rapid changes in amplitude though.

Switch effect? Switch effect from what device? The rectifying diode? That's the first part of the envelope detect. Switch effect from a negative feed back op-amp? Negative feedback is how you keep an op-amp linear. Switch effect from the BJT? A Bipolar Junction Transistor is a current amplifier. The base current gets multiplied by the Beta of the transistor to set the collector current until the collector to emitter voltage reaches saturation. (Not perfectly linearly, there is some compression before Vce saturates. But I'm not going into semiconductor theory 101 today!) What's left, maybe you think the resistors or capacitors will act as a switch.

Randall produced a very good, workable schematic. The only thing missing is the component values that a qualified or truly inquisitive person could figure out. If he doesn't have a second GA, I'll have to add mine to get him there.

redfred, that would be amazing if you could help find out what the values for the components would be. thanks

also, can you explain what this will do and how it will work? would i need to build 2 of these circuit? one for left audio and one for right audio? i only tought myself how to read schematic a few days ago (built a small amp to increase the output of headphones, i learned some from that. also it worked the first time!)

This is just a starting point: you need to do some guessing and tweaking when you know more about what's going on. First off your tapping the AC signal off a low impedance part of the circuit so the first 20 K resistor will ensure that you take almost no power from your "original" system.

Resistors left to right:

20K, 10K, 10K, 39K, 1K, 1K.

Capacitors left to right

1 µF 100 nF

In addition (why has no one pointed this out yet) you need another 20K resistor in parallel with the 1µf cap (otherwise the capacitor will just charge up and stay there.

Now the two 20Ks and 1µf cap give you a time constant of 10 milliseconds.

The 10 K on the +ve input to the op-amp doesn't (really) carry any current, so, could be almost anything.

The 39K and 10K in the negative feedback to the op-amp gives you a DC gain of 5 (nearly). This is the area where you'll need to mess about a lot. You might prefer to replace the 10K with a 2k resistor and 20K pot. in series.

The two 1K resistors will depend on your choice of power transistor which will itself depend on your choice of fan.

You'll probably want to use a 15 V power supply for a 12V fan (but if you can achieve the desired result with a fan running under speed then taking a 12V feed from one of drive supply cables in your PC would be OK. {You could also buy a replacement PC PSU to use as a 5V/12V bench supply: cheapest option if you're going to do a lot of this sort of thing}). Don't forget the power supply decoupling caps.: a 10µF electrolytic where the 15 V comes onto the board and a 100nF across the power rails at the op-amp should do.

I still prefer Redfred's idea but you'll learn more about electronics if you try to do this.

Good luck.

Randall,

I did notice the lack of a discharge resistor across the capacitor. I thought you were generating a simple peak hold circuit, anticipating that either the circuit board leakage, cheap capacitor, diode reverse bias current or amplifier bias current would be sufficient to bleed off the capacitor charge. To maintain the 20K impedance that the capacitor sees for your time constant and to maintain the same voltage level, the bleed resistor must be at least five times the input resistance of 20K. In other words, I would choose at least a 100K bleed resistor. Your calculation of 10 milliseconds for the rising time constant is correct for both resistors being 20K. The decaying time constant will now only be through one resistor and be 20 milliseconds, hence the concept of peak hold. Going with my suggestion of 20k input with a 100k bleed resistor will make a 19.6 millisecond rise (to a higher voltage) with a 100 millisecond decay time. (This should give the OP an idea of how to juggle these critical input components to get the desired effect.)

The input diode was not identified. Almost any switching diode, bipolar or Schottky will do for this job, but be careful not to accidentally choose a zener diode. A frequent diode I use for this are the 1n4446 aka 1n914a.

The classic 741 operational amplifier (Nat Semi LM741) will work but may require both positive and negative power to operate as anticipated. The Analog Devices OP290 and OP295 are dual (2 amplifiers per package, one for each audio channel) operate on a single supply. So the computer's +12V and ground supplies will work just fine. Which ever amplifier chosen, I recommend getting the DIP package for ease in assembly. Unless you wish to attempt surface mount parts.

The output transistor selection will be dominated by the anticipated maximum current draw of your fan (I_peak). Without knowing this parameter, no suggestions can be made. This transistor will likely have to be given a heat sink. The maximum heat dissipated by this transistor will occur when the fan is running at half speed. This maximum wattage will be (6V*I_peak)/2. But, there is another critical part missing from the diagram. An additional diode must be added, reverse biased across the collector and emitter of the output transistor. This diode will do the dynamic braking of the fan to slow the spinning down as the peak hold circuit request. Without this the back EMF from the fan can detonate the transistor. A small value resistor in series with this braking diode can reduce this effect but to protect the diode this must be included. The diode must be capable of handling I_peak. So the earlier signal diode will not be adequate. I recommend a power Schottky.

why not just run the fans with the 12vdc right from the computer power supply?

First, the "fan" comment seems to be a red herring or in less colloquial terms, an errant side comment. I believe the comment "fans" referred to using the DC voltage that power the fans to power the new audio amplifier.

Now to make your sounds louder, you have two options in audio. As suggested you could use a more powerful amplifier. This can be done with a DIY kit like http://parts-express.com/pe/showdetl.cfm?Partnumber=320-212 a 15W per channel, stereo amplifier or in the over kill category a kit from this company http://ampslab.com/ . If you still wanted an over kill that's not a kit you could try one of the myriad automotive audio amplifiers. Both of these approaches though require that you include a power supply (I don't recommend using your computer's power.) But you do run into diminishing returns for your speakers can only handle so much power before being damaged.

This leads to your second option, more efficient speakers. It is also quite likely that your quest for more sound has toasted one or more of your speakers. The combined suggestion somebody else posted of getting self powered speakers will probably be your best bet. In the over kill category, you might like these http://www.audioengineusa.com/a2_home.php

Randall, could you explain that to me again? i don't understand what your saying, sorry.

guest, i need an amp so that the motors speed will increase and decrease as the song gets louder or softer. the pc power supply will provide 12vdc yes, but i dose not change the voltage as my songs get loader

redfred, I'm not trying to make my speakers loader, i want my fans to turn faster. the reason I'm using an audio amp is because they put out more or less power as the music i play gets more bass or gets louder.

Ok now I understand your problem. Powering any of the above audio amplifiers with a separate power supply will likely solve your problem. For when you get a lot of bass, your computer's power supply and built in audio amplifier has to produce more power to drive your speakers. When these produce more power, more heat gets produced inside your computer chassis. Taking the amplifier and power supply out of your computer chassis will mean less power in the chassis, thus less heat. No need to change your fan speed.

If you insist on using your internal supply and internal amplifier (built in or kit) then you will have to modulate your chassis fan speed. You might consider this http://www.quietpcusa.com/Zalman-Multi-Fan-Speed-Controller-ZM-MFC2--P209C7.aspx or other less exotic fan speed controllers.

no, thats not it either. i don't need the fans for cooling purposes. I am using tiny 1.5' fans, I'm using them because they are small and brushless.

i then mount a mirror onto the fan andat an angle. i then shine a laser beam onto the mirror, this then makes a cirlce that or a laser tunnle if i use smoke in the room, but since im using a stero amp it can control 2 fans adding anoher "demension" to the laser beam. kind of like fan 1 = x axis fan 2 = y axis

If you use a speaker voice coil to tip a mirror back and forth instead of spinning a mirror you can get exact, repeatable and fast movement in one axis. Bounce the laser onto another mirror that tips in another axis and viola, laser spirograph.

Great idea: GA.

I would consider using a full wave voltage doubler circuit.

This will cause increased current draw from your amp so make sure it will handle the load.

Regards,

Keywalker

can someone explain what these two circuits Randall and keywalker have posted do?

how they would be used, etc.

GA:-

I was using the op-amp for DC gain as well as current gain, but that's a minor point.

For completeness:-

rectifier

DC amplifier and smoothing:-

And finally power amp

But Visible Noise,

I would seriously consider Redfred's suggestion of mounting tiny mirrors on the actual cones of some cheap speakers. You will probably need to "turn down the volume" on the speaker closest to the laser source, and get everything as close together as possible.

thanks!

if i mounted the cone of the speaker the design is more spirograph and does not make interesting spirograms like the fans with the mirrors.

Oh yes you will be able to produce spirograph patterns. Look at the Lissajous link. You will have to generate sums of various sinusoidal waves of different frequency, phase and amplitude to produce the patterns you want.

Frankly I don't see how a mirror mounted to a spinning fan will give you anything but a 60's era patrol car emergency light. But It is your project. Maybe you intend on other fixed mirror positions to translate motion. I also don't see how you can control the phase of your fan oscillations once you translate rotation to linear motion.

thats more work than needs to be done. I don't want to have to record/make sounds to create different designs, i just want the music to control it.

I have used this design with brushed motors before so I'm sure it works. I just need to get these fans to spin. The brushed motor caused too much vibration in the project box, this threw off the motors and just made squiggly lines on the surface

i will use a stepper motor design later as i learn more about electronics. I'm a freshman in high school and there is no class for small electronics, or any kind of electronics for that matter.

I don't think you quite understand what Redfred means:-

That dome in the middle of the cone already moves in and out sinusoidally (at the frequency of the music). Attach one side of a very small mirror to the centre of the cone, and anchor the other end of the mirror to a fixed point (on a framework). The laser pointer is now sweeping out a plane in space. Arrange the other mirror so that it moves the beam at right angles to the plane. Feed exactly the same signal to both speakers and you'll just get a plane diagonal to the first one. Delay the signal to one of the speakers and you'd get ellipses for a fixed frequency, but, delay the signal to one of the speakers by say 10 milliseconds with music and you'll get all sorts of great effects.

Good luck whatever you decide to do, and, come back here for more help on the electronics if you want.

Remember Wikipedia is your friend:- http://en.wikipedia.org/wiki/Operational_amplifier

Look half way down the article for basic inverting and non-inverting configurations.

i understand about gluing he mirror onto the speaker. but i don't get how i would be able to anchor the mirror. if the speaker only moves back and forth, how would i make the mirror turn left and right or up and down?

I like the idea about delaying the audio channelsthough.

You will probably not only get "vibrating mirrors" but also some pretty awful sound using speakers in this manner. Attaching mirrors may be problematic as well, at least as far as keeping them on there.

Hopefully the noise out of a pair of $5 speakers will be drowned out by the PA speakers. I think that trying to hold the speaker coil in the correct position after trying to remove parts of the cone would be a bit problematic: maybe he could glue some nylon "spokes" onto the original cones, then cut away the paper in between.

Attaching mirrors may be problematic as well, at least as far as keeping them on there.

I think he's up for the challenge. The secret is very small light mirrors.

Attach one side of a very small mirror to the centre of the cone, and anchor the other end of the mirror to a fixed point (on a framework).

Regardless of what type of audio output you have, the wave should look similar to what is pictured (taked from a screen shot of Audacity) at the right. This can be sent to the voltage doubler circuit just as well as the output from an audio transformer. The doubler circuit will produce DC with a voltage that is roughly twice the peak to peak voltage of the audio signal. This is exactly what you are wanting to send to the 12 volt fans. The louder the volume the higher the voltage. Also, you may have to experiment with capacitor values to be able to react to volume changes as you desire to. The biggest problem to me is that you are not matching your impedance with this fix which may cause you to blow your audio amp. Some sort of current limiting probably needs to be implimented which will again reduce your voltage available to the fans. For this reason an op amp circuit that controls a seperate voltage supply through the use of output transistors may be preferred but I wouldn't know for sure without trying the simple fix first. There are probably a lot of circuits related to monitoring volume and input levels with progressively lighting LEDs. Maybe one of these circuits would be useful to you. As the voltage rises more LEDs come on because of diodes being in the circuit in such a way as to allow this. These circuits are very responsive. I would try to keep it as simple as possible while trying to be as safe as possible though.

Regards,

Keywalker