super portable, super loud, long lasting, battery powered speakers

Step 2: Chosing the right combination of components

this step is to ensure that you don't go and buy components that are mis-matched.
the first thing that I found were the cones, I found these laying around my house. once you have your speakers open them up to see the cones themself, they should have a power rating and a resistance printed on the back, take note of these.

with the rating on the back of the speakers choose an appropriate amplifier chip, for me I had lot of 4 3.6W 4ohm cones, I decided to put them into two series connected sets, this gave me two satellite speakers each with a rating of 7.2W and 8ohms, the chip I found to match this was the TDA7057AQ, a quick search on farnell/digikey will find one to match your cones.

The amplifier chip will have a maximum input voltage in the data sheet, find the largest capacity battery you can that conforms to these voltage limits, I went with two 4 cell lipo batteries each with a capacity of 2250mAh wired in parallel to make a 4 cell pack with capacity 4500mAh

now you have all the major components worked out you can start the build.

Comments

[joshvonnieda]

Just a note if you really want super loud maybe upgrade on the speakers. I built something similar in a pelican case marine grade polk speakers and a rockford punch 30 amp. This thing is LOUD. Powered by a 20aH sla. Runs all night when fully charged. Wish i had your battery setup as mine is heavy.

[maxhysteria]

actually when you connect a pair of speakers in series, instead of raising the ohm load, like you said from 4 to 8, you will actually drop the ohm load, so it would be 2 ohms instead of 8. now if you dont believe me, just google it or look it up on youtube. if you were to have connected the speakers in parallel, you could potentially have even louder rave speakers that will last longer because less power is required to run them. Awesome project though man, this is probably the coolest boom box project ive seen on the website!

[jdread]

@maxhysteria, actualy you might want to google it... series wiring ADDS ohm load, not decrease. 4ohms x2 in series is 8ohms as he stated. Series wiring is positive to negative like batteries in a flashlight, stack them up then add them up.

[polop]

@maxhysteria indeed jdread is right, I think you have mixed a few facts up, series connected impedances add, also in general the lower the impedance connected to an amplifier the louder the speaker will be (more complicated than this really) but the power drawn will be proportional to the output power of the speakers, hence the batteries will run out quicker. Any way you can ask if you want a real analysis of the situation.

[T0C]

Why argue, if not sure just measure with a multimeter or other device. And another thing about the ohms, the higher they are, the lower the consumption and the smaller the heatsink has to be, if they are lower it's the opposite, higher power consumption, louder amp, bigger heatsink is needed...

[sconner1]

In series, the speaker impedances are added and the power use goes down because of Ohms Law.
In parallel, the impedance is half (as long as the two speakers are the same) and the power use goes up for the same reason.
The tradeoff is loudness vs. battery life.
Up to the point of the chip's and battery's minimum load and maximum current capabilities.
Too low of an impedance will:
If you're lucky, put the chip into protect mode
If you're not, fry the chip with too much current even with a proper heatsink OR
the battery won't be able to deliver current fast enough and cause "clipping" where the audio waves can't be reproduced properly and you get distortion.
Electric guitarists often want this effect but for music playback, it's ugly and causes ear fatigue.

[bengström]

When it comes to speakers it not that simple.
The resistance (impedance)varies depending on which frequency feeding the speaker, the resistance value is measured at 1kHz(?, not entirely sure).

[polop]

@TOC Don't worry this isn't an argument, just a discussion. Like Bengstr?m said there are more complicate forces at work and in general you should be absolutely sure what you are doing before you use a multimeter, in general an 8 ohm speaker has roughly 6.5ohm D.C resistance, but the impedance can only be properly described by the impedance curve of the particular speaker (impedance vs. frequency, can be done with a multimeter and a signal generator, but you might miss some important details) at resonance the impedance can be really quite high. But in the end none of this matters for this kind of instructable - it is meant to be quick and dirty way to get portable audio, using nominal impedance is fine for matching speakers to amplifiers. At some point I will upload a real instructable that goes into these details - leading to a truly high quality product.

[T0C]

I didn't say that when you measure it with the multimeter you would get an accurate reading, i was supposed to say "if you don't know it measure it with that and you would roughly get the impedance." also nice Instructable.


Btw i'm also working on a portable amp or two, and in about a week or so i'll post my design as well. (it isn't very big, but weighs about 2-3 KG with speakers that's it's only disadvantage... The good thing is it lasts about 1-2 days on full power non stop playing and it's 2x10 W ...)

[polop]

2-3 kg that is bretty good going, my new speakers weigh around 4kg, and have 72Wh of energy stored, or 3.5 hours at 2*10W.

I think you are likely to be missing some information some where (look at my very long post to someone about output power) according to you, I will be generous and 24 hours at 2*10W. this means you have 480Wh of stored energy. If you are using lead acid batteries that would weigh (480Wh) /(41wh/Kg) =11.7kg, that's a big battery.

If you used a good lithium ion battery (480Wh)/(128Wh/Kg)= 3.75 kg (this would be quite expensive)

That is just battery weight, plus cones and enclosure a speaker that could last even just one day at that power would weigh at least 5kg if you are incredibly clever about reducing weight. 2 days - obviously allot more than that. So you are likely over estimating the power you are actually using and probably the time that they will last as well. If your speakers do actually weigh 2-3kg and contain more than 480Wh of energy - call the patent office now - because you are going to be a rich person. ;p

[T0C]

Here's a short explanation:

That post was made 4 months ago, the amp from then is long dead (It could play about 17h or so on a full charge (on full power around 12h.)
The weight of the amp and speakers was >500g (good ol' AlNiCo magnet speakers) the rest of the weight was the battery (and case, tho the case couldn't have been more than 200g)
So, the 17h run was made with old cell-phone(look like cell phone batteries, but i don't think they are) batteries (nothing beats good ol' industrial batteries, the ones i used were 4.2V ~2110mAh (don't ask where i got them from) ) a total of 15 (...) batteries were used (and are still in use).
From what i calculated the brick was around 10Ah 12-13V
(Also, don't start calculating weight and stuff, these are not ordinary li-ion batteries they're Russian li-ion batteries, lol)

Also the weirdest thing was the batteries were(are) as light as a feather. If you took 3 batteries from these they would weigh as much as a BL-5C battery (aka the battery EVERY NOKIA phone uses today) don't have a good thing to compare to.


One last thing, the amp. I don't exactly remember it ( there were a lot of modifications to it), but from what i remember it was transistor based (when i say transistor based i mean it had a lot of bipolar and FETs) and was almost as big as the box, the output power was around 2x4-5W (10W speakers) it took 4 books and a lot of stupidity to make, but the end results were insane around 0.00042% THD made the sound crystal clear ....


That's all i remember
ask away if you have any questions.

[polop]

The no ordinary battery? if they are production batteries (cell phone) they will not have met your original specs any-day (even prototype batteries struggle to get over 200Wh/Kg) any way I am not trolling, but you shouldn't trust cell phone grade batteries ratings, especially if the source is dodgy. They are rated at very low draw and are overrated deliberately to outsell others (though I am sure the Russians are not as bad as the Chinese for that one, I have seen some Chinese products that break the laws of physics for only a few dollars :P)

About the amplifier - I would be interested in seeing the schematic/source of that - never have I seen a distortion figure that low for a practical audio amplifier. That is at least an order of magnitude better than ones I can find on the internet. Did you measure the distortion characteristics? or are you guessing ?

The case, less than 200g, was it made of cardboard? on my newest speakers my case weighs around 1Kg and it is too flimsy (partially my lack of skill I am sure, but I would think 1/5th of the weight would be horrific :P)

What size speakers were you using, 500g for 2 would be limiting you to some quite small ones, a decent single 4" driver is about 1kg (750g would be a little skimpy). if they are much smaller bass is going to go out of the window (especially if the box is not rigid enough).

anyway I would be interested in your answers from the sound of it you might have an interesting setup :)

[T0C]

About the battery, I just remembered they weren't from cell phones but were from a CPU(no idea what it means but its really cool)(Don't know the translation but it's a gigantic, not so old, CNC like machine).

The case is made from a plastic-like substance(a friend gave it to me, but it got lost about 1 month ago....damn...)

The speakers were small, yet powerful, not sure bout the dimensions and i think they were about 10W each... They're full range speakers with awesome highs and pretty clear lows (but when you run everything from one speaker sometimes the bass craps up the highs a little) also they're not that light one is about 300g, wasn't thinking when i wrote 500g overall.... pic related.


The amp, Actually me and a friend from Resprom(a Bulgarian audio company that made some of the best sounding amps and speakers I've ever heard.) developed/build the amp. He made most of the amp I just helped with the output stage(fun thing is after i made the output stage the dude gave me some weird looking transistors that turned out to be pretty fast/powerful) and filter.
Also he measured the THD with something like an oscilloscope.
All i remember from the schematic is....hmm.......well nothing.

Twas a great amp.

[karlpinturr]

If one speaker pair's values are different to/from another, would I go with a chip to match the smallest?

I assume I'd create all kinds of problems if I tried 'creating' similar values by wiring different values together in series/parrallel?

[polop]

oh and don't worry too much about the chip, as long as the rating is correct for the impedance - higher impedance is always ok - you need to think carefully when going under the rating (like 4 ohm in a amplifier designed for 8ohm) power rating is less of an issue than people think, if when you turn it loud and it starts to sound bad then it is too loud, because either the speakers are physical clipping, or the amplifier is - this is always bad. using a potentiometer that no body but you has access to is a good way of stopping people turning it too loud when yo are not looking.

[polop]

yeah - obviously matched pair is nice to have, and will likely sound the best, but if you are creative with the speakers you have you could make a nice set of speakers. for example if they are for boombox purposes (you want loud - and don't mind odd bits of sound going missing) you could have the larger and beefier speaker on one channel of the amplifier, and two smaller ones on the other channel in parallel. this would create a semi multi-way speaker (no flame i know it isn't a real one). then for the cabinet give chamber sizes that are proportionate to the speakers (large speaker has large volume etc.) and that would sound ok - and if you had a mono setting on your music player of choice you wouldn't lose any of the music. The larger speaker is likely to respond better to the lower portion of the frequencies and the smaller ones the higher, so this approach might give you an extended frequency response. but you are likely to have over powering highs, you could use a filter on the speakers to compensate - but by this point you may as well buy proper speaker drivers to save you the hassle of fiddling with the no- name mixed speakers. look at the site diyaudio.com - it as good advice on the real way to do this - I am active over there to along with people that will convince you to spend allot of money, but eventually you will end up with some excellent speakers that would make yo more happy than any store bought stuff.

[luckyvictor]

 so how do you come up with TDA7057AQ?? What is power rating exactly? does it have different meaning from component to components? let say the power rating of a resistor, because sometimes I see 1/2 watts or 1/4 watts resistors, what is the difference?

[T0C]

the differance is almost none. the 1/2 watt can support less current than the 1/4 watt (example a 1/2 resistor wold burn out at 1amp and a 1/4 at 2-3(no they wont it will be either less or more this,this is just an example))

[polop]

the inverse is true, 1/2W can support more current than 1/4. use P=(I^2)*R