New DIY-er needs "sound" power and build advice. (get it?)

Hello everyone,
First off, please excuse any layman's terms, dumb questions, and facepalms that might ensue from reading this post. This is my very first attempt in building any kind of speaker system. I'm a Mechanical Engineering student and life long music lover and audiophile. One day I heard of the Daimondboxx, but with the price, it's out of the question. Then I had the idea "why not build one for cheaper".
I have purchased a TPA3116D2 from Nobsound on Amazon, and have ordered the following drivers and crossover:
1 - Dayton Audio DCS205-4 8" Classic Subwoofer 4 Ohm
2 - FaitalPRO 4FE32 4" Neodymium Professional Full-Range Woofer 4 Ohm
2 - Dayton Audio ND25FA-4 1" Soft Dome Neodymium Tweeter
1 - Dayton Audio SD270-PR 10" Passive Radiator
2 - 5 kHz High Pass 4 Ohm Crossover
all but the amp are from parts express

What I have in mind is a AC/battery powered Bluetooth boombox in a sealed MDF enclosure. As far as the actual build, I feel good about being able to do it, BUT, as I've said above, the mechanical side is my thing, and the electrical aspects at this point in my studies is still something I'm shaky at best.
I'd like to have a woofer and tweeter on each L and R channel via the crossover, and the sub on the 100W channel.

There's a Youtube video titled "Build A Bluetooth Boombox Speaker (from Scratch)", and at time 11:30, he shows a wiring schematic.

MY QUESTION, is with the TPS3116D2, what kind of Battery, Power Adapter, and Charger should I be looking for. Also any other kind of "stupid newbie" mistakes I should be looking for. At this point, nothing is here yet, and I don't have anything else bought. What would you recommend for wattage and volt amount for this amp? I'm not apposed to ordering some 18650 batteries and making a pack. ( I can and have soldered) But how many to be able to power this amp without frying it and keep it powered for a decent time? What kind of charger for the batteries? Basically, HOW DO I POWER THIS THING CORRECTLY?

Thanks in advance for any answers or advice from someone who knows more about what I'm doing than I do.:)


2016-03-01 4:10 pm
On the seller's Amazon page, there is a hint that a laptop power supply might be a good choice for AC operation. That seems like a good suggestion to me - they are typically around 18V - 19V DC, and often capable of supplying 80W or more. That seems well suited to the (real) specs on this module.

For battery operation, I was going to suggest looking at 18650 lithium-ion batteries; but I see you got there before me! Some years ago, I used a number of home-made packs of these in electric-powered radio controlled aircraft models.

There are a number of safety considerations with all batteries, lithium ones in particular. A major one: whatever charger you use *must* be designed specifically to work with the particular chemistry of lithium battery that you buy - there are slight differences in fully-charged voltage, which can translate to ruptured batteries, or worse, high-temperature flames, if the wrong charger is used on the wrong battery. :eek:

The 18650 cells I used came out of DeWalt cordless power tool packs. They were a LiFePO4 chemistry, fully charged at about 3.6 volts per cell, compared to about 4.2 volts per cell for more common lithium-polymer packs used in phones, tablets, etc.

At that time, the best chargers were specifically designed for the radio-control market. They had the best prices, and the best power capability. (You need lots of power to fill up these high-capacity lithium packs in a reasonable amount of time.)

These RC chargers usually required a number of manual steps before they would charge a pack. The charger has to "know" how many cells are in the pack to avoid applying too much voltage, and it has to "know" the cell capacity (Ah) to determine the maximum charge current that's safe. So you had to punch these in manually before the charger would start.

I don't know if 18650 charger availability has changed in the last few years - certainly these cells are much more widely in use now, so there may be many more charger options. But, once again, make 100% sure that the charger is compatible with the specific cell chemistry.

To be clear: I'm suggesting using a surplus or thrift-store laptop power brick for AC operation, but NOT for charging the lithium battery pack. Wire the AC adapter so the batteries are disconnected when you run off AC. Use a separate charger designed for these specific lithium cells to charge them.

Now to the specs on the module. On a lot of these modules, advertised power is pure fiction. Fortunately, a little engineering math can reveal the truth.

So: a good way to find the approximate power you can get from a NON-BRIDGED amp, at a specific battery voltage (Vb, say) and speaker impedance (R) is this:
1) Subtract 3 from Vb. This is a guesstimate to allow for saturation losses in the amplifier circuit.
2) Square the result.
3) Divide the answer by eight times your speaker resistance.

Applying this to an 18V laptop power supply:
Subtract 3V gets us to 15V.
Square 15V, gets us to 225V^2
Eight times 4 ohms = 32 ohms
RMS power = (225 V^2/ 32 ohms) = 7 watts (RMS).

I looked up the TPA3116, and each chip actually contains four power amps; they are designed to use a pair in bridge mode for each channel.

In bridge mode, an amp can theoretically deliver 4x the power of non-bridged operation; so the previous 7 watts RMS turns into roughly 28 watts, RMS, for each stereo channel.

This is an approximate calculation, but will get you in the ballpark. With a good 18V power supply, then, you can expect somewhere around 30 W RMS per channel for the left and right channels (using 4 ohm speakers).

What about that ".1" subwoofer channel? As far as I can tell from the meagre information on the sellers Amazon page, it consists of two of these bridge-mode amps in parallel. IF you were to hook up a 2 ohm subwoofer, it would deliver roughly 56 watts RMS with a very stiff 18W power supply.

The catch being, you have a 4 ohm subwoofer, not a 2 ohm one. Therefore the amp, paralleled or not, can only deliver the same 28W or so into the sub, as it does into the L and R speaker channels.

This may not be much of a catch in practice - you may not need all these watts to achieve the volume you want.

So: if we estimate 30W each for L, R, and subwoofer, that's around 90W RMS total output. Even with very efficient class D amps, that's going to need well over 100W DC to power it fully. Ideally a laptop brick that can supply 120W or more - but I'm not sure such a thing would be easy to find.

On the battery side of things: a single 18650 cell puts out about 3.6 V hot off the charger, about 3.3 volts unloaded after a few minutes of settling, and a bit less under load, depending on how heavy the current draw is.

That suggests a 5-cell pack (18V hot off the charger, maybe 15V under load), or a 6-cell pack (21.6V hot off the charger, maybe 18V under load). I think the 6-cell pack is getting dangerously close to the maximum supply voltage of the amp module, but I'm conservative about these things. I don't like the smell of overcooked electronics...

You can also make estimates of how long the pack will power the amp - but the actual time will vary wildly with your volume setting.

To give you a very rough estimate: say you use 3.4Ah cells, and make a pack out of 5 of them.

We'll estimate pack voltage under load at 15V, so this pack can deliver 15V x 3.4 A for 1 hour; that's 51 watts for an hour.

A crude estimate, then, is that it will deliver twice that power for half that time; 102 watts for 30 minutes.

102 watts is fairly close to our estimated power demands for your amp with all three channels cranked to maximum power - so we can say with some confidence that a 5 cell, 3.4 Ah, 18650 pack could power your amp for roughly 30 minutes if you were blasting full-power continuous sine waves out of all three speakers, 90 watts of unbearable, ear-destroying, loudness.

If, on the other hand, you turned the thing down to a mere 9 watts total (that can still be very loud), the same battery should be able to power the amp for ten times as long: that's five hours of continuous operation.

You can use a very similar calculation to estimate how much power your lithium battery charger needs to be capable of delivering: a 50 watt charger would be able to charge the pack fully in about an hour. A 10-watt charger would take five hours. (In practice, both chargers will take a lot longer than that - the last part of the charging cycle involves low currents and lots of charging time, unfortunately.)

So, if you like these numbers, I think we have the basis of a plan. If I was in your shoes, this is the route I'd take.

But I'm sure other smart and knowledgeable people will chime in, too - so let's wait, and see what they suggest!

Thank You For the input!!
It's rated at 24V (the amp card), and I don't want to come close to this. One of the things I'm trying to keep in mind is to have a bit of an overbuild for safety factor's sake, but I do want it to be able to kick out some serious sound when circumstances call for such.
Here's a link to the actual one I ordered. Apparently they "upgraded" from their previous version of the 3116. Nobsound®Latest upgrated 2.1 digital power amplifier board TPA3116D2 50W+50W+100W 12V-24V: Electronics

I'm assuming if I do the 18650 pack route, the higher capacity of the batteries themselves, (the miliamp rating, mah) that will increase the playback time on a single charge? I see Ultrafire has a 3.7V 18650 with a 6000mah rating. As far as charger goes, I'm assuming I will need a charger installed that is spec'ed to charge all the batteries, not just one? And yes, I will make sure to get a Li-ion specific charger. I'm also thinking I'll have a 3-way switch, one switched for AC power, one for 18650 power, and one for the power off all together. I suppose I could turn down the gain for the L and R channels to kinda balance out the sub being 4ohms. Save a bit on battery power too.

Again, thanks for any and all help!!!


2016-03-01 4:10 pm
Thank You For the input!!
It's rated at 24V (the amp card), and I don't want to come close to this.
You're welcome!

I have the same approach, best to leave a bit of a safety factor. Things fail more frequently when pushed to their maximum ratings, whether they're electronic or not.

I do want it to be able to kick out some serious sound when circumstances call for such.
Why not start off powering the thing only from AC, and make sure the module meets your loudness requirements?

If the module holds up at full power, that is a *lot* of power, IMO, particularly if you will be listening to popular music with a restricted dynamic range.

If, on the other hand, your favourite piece is Tchaikovsky's 1812 Overture, complete with cannon blasts, then you may have a problem. :)
Here's a link to the actual one I ordered.
That's the one I found via Google, and that's the chip I looked up the data sheet for.

Data sheet: TPA3116D2 | Analog Input Class-D | High-Power Audio Amplifiers (>50W) | Description & parametrics

I'm assuming if I do the 18650 pack route, the higher capacity of the batteries themselves, (the miliamp rating, mah) that will increase the playback time on a single charge?
Yes, you're quite correct. Max playback time will be roughly proportional to the pack size. If you use 6 Ah cells instead of the hypothetical 3.4 Ah cells I used in the example, you can simply multiply all duration estimates by 6/3.4, or about 1.75.

I didn't mention it before, but rechargable cells have a number of ratings - "C rate", which tells you the maximum safe discharge rate, is one. There is also a maximum safe charge rate, which is much lower than the max discharge rate. (This is one of the things Tesla and other EV manufacturers struggle with.)

As far as charger goes, I'm assuming I will need a charger installed that is spec'ed to charge all the batteries, not just one?
You'll want a charger that can cope with a 5-cell 18650 pack (6 cell if you go that route), yes.

The RC chargers I mentioned are flexible: they will have an upper limit, a maximum possible number of series cells, but up to that maximum, you can enter the number of cells manually. So the same charger can handle, say, a 3-cell battery, or a 5-cell one.

Cordless tool manufacturers will take a cheaper and safer approach, manufacturing chargers specific to one size of battery pack. The customer can't accidentally punch in the wrong number of cells (which can start a fire and burn your house down, no kidding).

And yes, I will make sure to get a Li-ion specific charger.
Be careful about this, regular Lipo/Li-ion cells need to be charged up to 4.2 volts. The LiFePO4 cells used by DeWalt (and originally developed in the USA by A123) need to be charged to a maximum voltage of 3.6 volts. Beyond that, either the cells will be damaged, or you will have a high-temperature fire to deal with. :eek:

So a regular lipo/li-ion charger is NOT suited for these LiFePO4 cells. You want a charger that is specifically designed for this (different) cell chemistry.

I'm also thinking I'll have a 3-way switch, one switched for AC power, one for 18650 power, and one for the power off all together. I suppose I could turn down the gain for the L and R channels to kinda balance out the sub being 4ohms. Save a bit on battery power too.
You definitely want the "power off altogether" position, so the batteries don't get drained totally dry while the speaker is sitting there idle during exam week!

You could certainly to the 3-way route you suggested, or you can use a couple of simple rectifier diodes to isolate the batteries from the AC power supply. You insert one diode in series between, let's say, positive lead of the battery pack and the amp circuitry. The second diode goes in between the DC power jack from the laptop brick (positive lead, again) and the amp circuitry.

What this does is prevent the battery pack from discharging into the power brick, and even more importantly, prevent the power brick from overcharging the lithium pack and starting a fire. If AC power is present, the batteries will effectively be disconnected, since the voltage at the amp board will be a little higher than the battery pack, so the series diode will be reverse-biased.

I am not kidding about the fire issue - not only can lithium batteries store a lot of energy, which is always dangerous, but some lithium chemistries are prone to burst into extremely high-temperature flames if the cells aren't treated with kid gloves. The LiFePO4 chemistry was invented partly to reduce this risk, as I understand it; the A123 cells came from research initially done at MIT, if I recall right.



2016-03-01 4:10 pm
That will probably do the job (though it may not power everything up all the way to maximum continuous power).

I think you can get away with spending a lot less, though. These sorts of laptop chargers have been sold by the tens of millions, and for a decade or two now. That means there are millions of surplus chargers out there. I've found them for $5 at local thrift stores.

You may have to deal with changing the plug on the end of the wire, if it doesn't match your amp module - but, clearly, you can already solder, or you wouldn't be tackling this project!

Thank you very much Gnobuddy!
I talked with one of my physics professors who agreed to double check everything, so I don't blow myself up. :) The good Dr. built his own system before, but not battery powered or bluetooth, but he also teaches some electrical classes, so I'm sure he's qualified.

Thanks again for pointing me in the right direction!
- JF
I also found this as a possible solution to my battery issue. I think this should be able to deliver the performance and battery life I'm looking for. It's kinda at the top of the budget, but I want to do it right, and have it do the things I want it to the first time. The RC stuff is kinda universal, so I should be able to find a charger for this and hook it up? I think? I just need to make sure it's sufficient to charge this pack.

What ya'll think? FLOUREON 6S1P 22.2V 10000mAh 25C Li-Polymer Lipo RC Battery Pack with Deans Plug for RC Helicopter RC Airplane RC Hobby: Toys & Games


2016-03-01 4:10 pm
You can certainly find an RC charger for that pack (though it will be expensive).

I'm sure you can use the pack, and it will work, and let your project run for an entirely ridiculously large amount of time, on batteries alone.

But: personally, I would NOT, repeat NOT, use that pack in your project. It's a safety issue: this is a regular lithium-polymer pack. If you overcharge it a bit, metallic lithium plates out of the internal electrolyte, onto the internal electrodes. Next thing you know, you have a huge, powerful, firework on your hands. High temperature flames, toxic smoke, mysterious stinky goo that is very unlikely to be good for you. If you're unlucky, you may also burn your house/ car/ garage down.

That's not all. If you over-discharge it a bit, you can damage the pack. Next time you charge it, it can puff up, and once again turn into a powerful incendiary device.

There's more. If it suffers a physical impact, that can cause internal damage, which in turn can cause the incendiary device to show up.

Still more: leave it in a hot car in summer, and if the ambient temperature climbs high enough, lipo packs can spontaneously combust. The resulting fire is hot enough to total a car in minutes.

I'm not kidding. This scary stuff is real. A decade or so ago, in the early days of lipo use by RC hobbyists, there were a frighteningly large number of people who experienced lipo fires.

Over time, the chargers got safer, and some RC hobbyists wised up and began to take extreme precautions. But those who don't, usually pay for it.

I personally knew *THREE* RC hobbyists people who were not sufficiently rigorous in following all the safety rules, and who paid for it with lipo fires. All three fires were inside vehicles (these clowns were charging their lipo packs inside their cars, unsupervised, while they were at the RC flying field, flying their models).

Two of the vehicles were completely totaled. One was new, had full coverage, and the insurance adjusters didn't realize how the fire had started; that guy got a new replacement vehicle. The other guy lost his only vehicle (a Ford Explorer, IIRC).

The third guy had his fire inside a Toyota pickup truck. The entire passenger side area was torched to cinders. The plastic dash melted and warped. Amazingly enough, the controls and drivers seat survived, and the truck remained drivable, though it reeked of nasty chemicals inside the cab.

Cars are expensive enough, but some people have lost houses or garages or both to lipo fires. And house fires can kill people; far worse than the loss of any material object.

If you Google something like "lipo fire burns house", you'll find lots of scary photos and videos.

I saw plenty more lipo fires at the flying field. One pack burst into flames when the model it was in crashed into the ground (pilot error). One burst into flames because a bad soldering job caused a dead short across the battery terminals. Others, I don't remember the details any more, other than the smoke, the smell, and the burned grass around the packs.

One of the big selling points for the LiFePO4 batteries developed at MIT and sold under the brand name A123 was the fact that they didn't burn in the same way. I'm not a chemist, and cannot speak with any authority on the reasons, but the altered chemistry and construction meant that these batteries did not turn into flaming fireworks if things went even a little wrong. I hope those qualities still apply to the current (non-A123) LiFePO4 cells and packs.

Full disclosure: I myself used lipo batteries in my own RC planes for years. I also did intensive research to find out how to remain safe, and I never let my guard down. I transported my lipo packs in fire-proof containers with Ziplock bags full of sand placed around and over them. I only charged my packs outdoors, and I charged them inside concrete breeze-blocks, sitting on a concrete paver, with a large Ziplock bag full of sand placed on top. I always stored packs half-charged, in the 'fridge. I inspected all my packs regularly for signs of puffing or damage. (I caught several that had begun to show puffing, and disposed of them by discharging them, immersing them in salt water, and puncturing them to let the salt neutralize the internal electrolytes).

The idea of the sand-in-a-Ziplock-bag is that if there was a fire, the plastic bag would melt, and the sand would smother and contain the fire. With less chance of burning fragments spraying in every direction, there was less chance of setting something else nearby on fire.

With all those precautions, I never experienced a lipo fire myself. But, I treated these things as though they were loaded guns. Those of my RC acquaintances who weren't equally careful, experienced lipo fires or scary near-misses.

So: don't do it, please don't do it. A large lipo pack like that in long-term use, probably hidden out of sight inside an enclosure, is just too big a safety risk. (Do you know about the Boeing 787 Dreamliner lithium battery fires? If not, Google!)

There is no such thing as a completely safe battery (stored energy is always dangerous). But the LiFePO4 packs are safe enough for cordless tool companies to put them in mass-production consumer products, as long as they're treated and charged properly. So the risk is much lower than using a large RC lipo pack.

Speaking of LiFePO4 packs: if you solder them up yourself, remember not to let the metal parts of the soldering iron accidentally bridge across nearby cell terminals! One of my friends did that by mistake - two 2100 mAh A123 cells put out such a big spark that it melted a hole in the side of the metal tube that holds the soldering bit in his Hakko soldering iron!

Just in case I didn't say it enough times already: PLEASE DON'T USE A LIPO PACK!



2016-03-01 4:10 pm
I almost forgot: a 6 cell lipo pack comes off the charger with 25.2 volts terminal voltage (4.2 volts per cell). That is too much for the amp modules you're using.

Don't use any lipo pack, and in particular, don't use a 6S (series) lipo pack!

(LiFePO4 cells have a lower voltage per cell, as I mentioned in an earlier post.)

(and the face palm I more than likely just caused)

I will NOT be going with the LiPO option. Jesus... Sounds a little like cheap thermite, something I don't want to be around when it's reacting. I'll look for a Li-ion option again. I'm just kinda disappointed with the capacity results.

I'm also getting conflicting information as far as what a li-ion pack is claiming to be able to do. Let's say you have 6 3.6V 3200mAh batteries in a pack, I understand the voltages adding up, but I was under the assumption that the mAh was not added, and that the capacity rating of one can be regarded as the entire cell, as the entire cell is being drained. Maybe I'm overbuilding though.

In any case, Thank You again for the huge heads-up!!!
- JF
Serial and Parallel Battery Configurations and Information

SO... with the proper configuration, I am able to increase the mAh ratting, without stepping up the voltage by parallel configuration, and then step up the Voltage to what's required with a series configuration of the paralleled batteries. Or vice versa I suppose.

Thought I'd share the findings.
Again, Thanks Again for making sure I don't melt my face off with a chemical explosion.
- JF


2016-03-01 4:10 pm
I'm glad you decided against the lipo pack!

Yes, you can parallel packs to get larger capacity (mAh).

My suggestion: first build, test, and debug the circuit on AC power only.

Then try it on battery power, without using paralleled packs. You may very well find that you have all the duration you actually need. (There is at least one quite expensive commercial portable guitar amp that runs on just four AA cells, and claims to run for hours.)

The reason for this is that, most of the time, you will be using less than 5 watts, averaged RMS power, to listen to music - and that will be very loud, unacceptable in most apartments or condos.

People sell 100-watt amps cheap now, because the technology allows them to be made, and bigger numbers attract more attention. But that is the sort of power that causes loudness-induced hearing damage on a dance floor, never mind a typical single-family living room.

If you then find you really do need more battery capacity, you can always parallel a second pack (same number of series cells) to double the pack capacity.

I built a 2.2 kilo RC aircraft that would fly for most of an hour on ten 2300 mAh LiFePO4 cells - and model aircraft tend to use a lot of power. I never used that many cells again after that, usually just 5 cells.