# Audio amplifier project - help with speaker ratings/types

#### jameselder232

Joined Oct 31, 2010
40
Hi, my task is to build an audio amplifier for an mp3 player 3.5mm output, the spec says to run it on 3 AA batteries (4.5V), and to have a pre-amp, tone control and output amp.

Now i have the circuit all wired up and working, however we are given these cheap miniature loudspeakers to use (mylar) and the problem is they are terrible at handling bass, it just sounds like noise really, however if i remove the bass with my bass control circuit i get a nice output. I'm sure if my research is right they are only designed to handle frequencies above 400hz which is obviously the problem.

Now we are allowed to purchase other components if they are cheap and readily available, but i would like some advice on purchasing some speakers that can handle the full audio range (i think it was 20hz-20khz), which are small in size and power rating, and are cheap and readily available. We are currently using speakers i believe are rated at 8Ω with a power of 1W max.

As you can tell i don't really have any experience when it comes to loudspeakers and audio, so i don't know what info you might need or how stupid this question may be.

Thanks,
James

#### jameselder232

Joined Oct 31, 2010
40
Is it right to assume bridging an inverted output to the non-inverted output will increase the power output and the volume?

Also by connecting the speakers in parallel i reduce the resistance and hence increase the volume? Does this affect the power output also then?

I understand from research that obviously i have to be careful and make a trade-off between volume and speaker rating.

I believe my circuits maximum output power is 0.1W currently. This is only according to simulation though.

And from research i believe the speaker we might be using is the 388:

Impedance 8Ω
Power 0.1W
f0 (Hz) 550±100
f(range) f0-3500hz

which therefore means the speaker has too low a rating for my circuit, and so i definitely need a new speaker with a higher power rating and a larger frequency range. But it needs to cost £3 or less, and be available from a big component supplier such as rs or maplin. I'm from the UK. It also needs to have a diameter of ~40mm.

Last edited:

#### R!f@@

Joined Apr 2, 2009
9,682
Listen buddy.
If u want to produce good bass, then a 1W amp and 4" speaker is out of the question.

Bass needs lot's of power and a large volume of air movement, which means powerful speakers design for low range

#### jameselder232

Joined Oct 31, 2010
40
Listen buddy.
If u want to produce good bass, then a 1W amp and 4" speaker is out of the question.

Bass needs lot's of power and a large volume of air movement, which means powerful speakers design for low range
Okay but i would just like to be able to hear the bass more clearly, because it just sounds muffled at the moment.

Do you think i would be able to achieve this with my setup? I mean I'm sure the speaker is a big factor on this at the moment. It's definitely not designed for the bass frequency range, and the power rating is too low, so that's a good start i guess. But i can only work with the specs I'm given so I'm never going to get great bass, but i think i should be able to improve the clarity a bit.

How about the questions i asked with regards to the power output/volume output?

#### R!f@@

Joined Apr 2, 2009
9,682
I think U need to get full range speaker.
There is no way to get descent bass from small ones. But U can get around 300Hz.
and the high will too be cutoff at around 10K.

A single speaker is not an option for good clarity audio reproduction.

There are speaker like the ones used in TV's or altec lansing cubes or like boss cubes. these can produce descent sound if driven without clipping.
But I doubt they are at 3 bucks.

#### jameselder232

Joined Oct 31, 2010
40
I think U need to get full range speaker.
There is no way to get descent bass from small ones. But U can get around 300Hz.
and the high will too be cutoff at around 10K.

A single speaker is not an option for good clarity audio reproduction.

There are speaker like the ones used in TV's or altec lansing cubes or like boss cubes. these can produce descent sound if driven without clipping.
But I doubt they are at 3 bucks.
okay, well i just found this speaker:

http://uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=4318484

2in, 2W, 8Ω. With a frequency response between 180Hz-17kHz at £3.70, so its looking like a good option at the moment. We are actually using a pair of speakers on the output for the final product, but that won't really help the output sound will it? it'll just enable us to separate different frequency ranges onto different speakers (if we choose to). But obviously it means we can vary the load impedance.

I have got my circuit to produce a decent quality sound without clipping so far, so hopefully a good speaker will make a big difference. It's gotta be better than with those cheap miniature 0.1W speakers we are given

Last edited:

#### R!f@@

Joined Apr 2, 2009
9,682
yes tht wud do fine for ur needs.

Remember to drive it a bit higher than rated power. This will help prevent clipping and clean sound enuf for the speaker.

What u need is around 1.5Watts/ channel

#### jameselder232

Joined Oct 31, 2010
40
So your saying the power into the speaker should be higher than the power rating of the speaker? and this will prevent clipping? if you don't mind explaining, how come this is the case?

i thought the idea was to run it at the rating of the speaker, but the speaker has a little give for fluctuations anyway. Also I'm not sure if i can get 1W from my circuit, i mean i have a gain of 2 on the preamp and a current gain of 1000 on the output. I can't increase my preamp gain as the rails are at ±2.25V due to the spec being 4.5V and the maximum input voltage being 2V pk-pk, and so there will be clipping on anything higher. And my simulation software shows a peak power output of 0.2W, if i'm reading it all correctly, with the average being around 0.17W.

Its really difficult to build this circuit when the range of input voltages i have found is so large (0.2-2v pk-pk).

#### R!f@@

Joined Apr 2, 2009
9,682
Clipping causes distortion and horrible sound.

And clipping is the primary cause for speaker burnouts.

By running a speaker with an amp higher than it's power gives the amp a margin to keep it cool and ample space to avoid clipping at full volume that the speaker can handle.

So to say if u max the volume ( if power out is set accurately to it's respective speaker ) the speaker will operate at it's peak with no distortion and the amp will run cool for extended periods.

#### jameselder232

Joined Oct 31, 2010
40
Ah sorry i misinterpreted what you were saying earlier and i misquoted the power rating of the speaker.

I thought when you said drive it higher than rated power you meant make the power into the speaker greater than the speakers rated power, but i realised now 'drive' refers to the load power rating.

And i was thinking the speaker had a power rating of 1W but its not its 2W, and so obviously 1.5W is slightly lower and gives room for fluctuation, keeping it running cool.

However I'm still not sure if i can even output anything near to 1.5W with my circuit. But if i'm right and i can increase the power by bridging an inverted output to the circuit i should get up 0.4w, but i still need to get this higher...

I will carry on with the research and testing and see how i do but i do believe I've maxed my circuits power output when using 4.5V supply and an input range of 200mV-2V pk-pk.

Thank you

#### R!f@@

Joined Apr 2, 2009
9,682
What is the amperage of ur 4.5V supply

#### jameselder232

Joined Oct 31, 2010
40
What is the amperage of ur 4.5V supply
I'm not sure but i'll measure it today, only problem is that in labs i only have a DC power supply to mimic the 3 AA batteries, i don't actually have the batteries to test with, so i would expect there will be a slight difference.

#### R!f@@

Joined Apr 2, 2009
9,682
So in reality u will use 3 AA batteries.

Are they 2200mA or 2500mA Types?

#### Audioguru

Joined Dec 20, 2007
11,251
The cheap little 2" speaker has a flat frequency response from 300Hz (no bass) to about 2kHz. The low end drops off below 300Hz and the high end peaks at 3.5kHz then notches at 5kHz and goes all over the place at higher frequencies. It will sound pretty bad (cheap).
The bass will actually be worse because an enclosure is needed which increases the resonant frequency which reduces bass frequencies even more.

If you have a 4.5V battery as a power supply then it might actually produce only 3.5V to 4.0V.
If it produces 4.0V then a bridged amplifier will have a peak-to-peak output voltage into an 8 ohm speaker of only 4V at the most which calculates into power in the speaker of only 0.25W at clipping.

#### Audioguru

Joined Dec 20, 2007
11,251
"Peak" power is phoney-baloney power with a number that is double the real RMS power.

#### jameselder232

Joined Oct 31, 2010
40
Hi, attached is the circuit i am using.

The power supply on the first op-amp is just for simulation purposes as I'm using 741s, and since i don't amplify above 4.5V with my feedback resistors, it doesn't really matter for the simulation purposes.

It's a work in progress so any advice would be great =) I still have to do impedance matching and add some decoupling capacitors. These are circuits from the internet which i am cascading together. The first part being the pre-amp, the second being the bass control, third the treble control and lastly the output amp.

With the AA batteries, this is for a project and the only spec I have is to use 3 AA batteries, as far as which type of AA batteries, that hasn't been specified and I'm not sure what would be best.

With regards to the speaker choice, I'm only allowed to use cheap speakers, unless i decide to buy them myself, so i can't really do much there.

Thanks

#### Attachments

• 36.8 KB Views: 50
Last edited:

#### Audioguru

Joined Dec 20, 2007
11,251

The 741 opamp is 42 years old and is horrible for audio.
It has a high amount of noise (hiss).
Its slew rate is low causing trouble above only 9kHz but audio goes to 20kHz.
It is made for a 30V total supply and few will work with a supply that is less than 10V.
Use an audio opamp like an OPA134 that works with a supply as low as 5V.
An OPA4134 has 4 audio opamps.

I simulated your power amplifier. I added a series input resistor to provide negative feedback for reduced distortion. I adjusted the biasing resistor for symmetrical clipping.
The resulting amplifier has about 10% distortion and also has crossover distortion. Its output power at clipping is 94mW RMS using "typical" transistors and will be much less power if weaker transistors are used. Headphones sitting on a table might be louder than your amplifier driving a speaker.

#### Attachments

• 27.3 KB Views: 54

#### jameselder232

Joined Oct 31, 2010
40

The 741 opamp is 42 years old and is horrible for audio.
It has a high amount of noise (hiss).
Its slew rate is low causing trouble above only 9kHz but audio goes to 20kHz.
It is made for a 30V total supply and few will work with a supply that is less than 10V.
Use an audio opamp like an OPA134 that works with a supply as low as 5V.
An OPA4134 has 4 audio opamps.

I simulated your power amplifier. I added a series input resistor to provide negative feedback for reduced distortion. I adjusted the biasing resistor for symmetrical clipping.
The resulting amplifier has about 10% distortion and also has crossover distortion. Its output power at clipping is 94mW RMS using "typical" transistors and will be much less power if weaker transistors are used. Headphones sitting on a table might be louder than your amplifier driving a speaker.
Okay well I'm limited to using the op-amps available at my labs, namely the 741 or the 358 or the TL074. I was thinking to use the 358 instead? With your simulation did you use the 741s or one of the audio op-amps?

I expect some crossover distortion as it is a class AB amplifier, but are you telling me there's quite alot of crossover distortion here for a class AB?

And lastly i have actually attached that above circuit to the cheap 8Ω speaker we are provided with and it is louder than earphones for definite, but not particularly loud for an amplified speaker, and so I'm looking into increasing the power output by bridging an inverted output, and doing some impedance matching.

But the main problem at the moment is to clean up the distortion, and also to try and buy a better speaker as the current one is awful at handling bass frequencies. Then hopefully i can bridge the inverted output to get more volume, and it should meet the spec anyway.

We are thrown into these projects with little experience in audio and also transistors, and so I'm just trying to do the best I can. Thanks for the advice, I have some idea about what i need to do next now

#### marshallf3

Joined Jul 26, 2010
2,358
Great speakers at inexpensive pricing: (sounds like an ad, but it isn't)

http://www.partsexpress.com

Another good source would be car audio speakers, some really good 5-1/4", 6-1/2" or 6"x9"can be found used on eBay or at a local car audio store leftover from someone's upgrade.

Remember - to get bass there needs to be an enclosure, and a properly ported one increases efficiency.

#### Audioguru

Joined Dec 20, 2007
11,251
Okay well I'm limited to using the op-amps available at my labs, namely the 741 or the 358 or the TL074. I was thinking to use the 358 instead? With your simulation did you use the 741s or one of the audio op-amps?
I have used TL071, TL072 and TL074 in most of my many audio circuits but with a supply voltage that is higher than their 7V minimum. They are low noise and have a full output bandwidth to 100kHz. They are class-AB and have distortion that is only 0.003%.

I expect some crossover distortion as it is a class AB amplifier, but are you telling me there's quite alot of crossover distortion here for a class AB?
Class-B circuits have crossover distortion. Most opamps and most audio power amplifiers have a class-AB outputs and distortion that is almost zero. The OPA134 has distortion of only 0.00008%. Most good audio amplifier ICs like the LM3886 have class-AB outputs, no crossover distortion and total distortion at full output of typically only 0.003% at lower frequencies. The distortion increases at higher frequencies to 0.015% but you can't hear those very high frequency harmonics anyway.

And lastly i have actually attached that above circuit to the cheap 8Ω speaker we are provided with and it is louder than earphones for definite, but not particularly loud for an amplified speaker, and so I'm looking into increasing the power output by bridging an inverted output, and doing some impedance matching.
Your amplifier schematic does not have enough output current to be used as a bridged amplifier because its current will be double the current of your single-ended amplifier. Most ordinary car radio amplifier ICs like the TDA7240A are bridged for about 3.5 times more output power than a single-ended amplifier.

Vacuum tube amplifiers matched their output impedance to the speaker impedance, but then the speaker had poor damping of its resonances. Half the output power was wasted by matching impedances.
Solid state amplifiers never match the speaker impedance, instead they use high internal gain and lots of negative feedback for extremely low distortion and an output impedance of 0.04 ohms or less for excellent damping of resonances in speakers.

But the main problem at the moment is to clean up the distortion, and also to try and buy a better speaker as the current one is awful at handling bass frequencies. Then hopefully i can bridge the inverted output to get more volume, and it should meet the spec anyway.
A transistor is frequently used in amplifiers to replace the two diodes. Its voltage drop can be adjusted so that crossover distortion is eliminated but the idle current in the output transistors is low.
Amplifiers do not have just 3 transistors. You should do some research for 4 and 5 transistor amplifiers that have very low distortion and maybe have enough output current so that they can be bridged.

I recently bought a brand new speaker (donated by a store) at a Thrift Store for $3.00. It is a 4" full-range driver in a ported wood enclosure. I drive it with an LM390 amplifier IC with 1W for the TV in my computer room. It is pretty loud with 1W. It has a response of 70Hz to 15kHz. Its bass is pretty good for a cheap little speaker and it might have sold for$30.00.

My computer speakers have 3" drivers with big magnets. Each has an IC amplifier with an output of 3.5W. They have a response of 80Hz to 15kHz.

EDIT: I forgot. The lousy old LM358 dual opamp (an LM324 quad opamp is the same) is one of the first low power opamps. It had its power reduced by eliminating the biasing of its output transistors so its crossover distortion is 3% and its bandwidth stops at 2kHz.

#### Attachments

• 15.2 KB Views: 39
Last edited: