|
|
| Elliott Sound Products | Project 169 |
I'm not about to publish the URL, but there is a seriously demented power amp available from some dubious characters in Germany that costs a 'mere' 800-odd Euros, and is basically nothing more than a car amplifier IC mounted onto a plank of wood. It's supposed to be 'magical', but in reality it's simply yet another attempt to separate (gullible) people from their money, with the promise of audio 'nirvana'.
As you probably guessed from the title, this amp is intended to be powered from a car battery, with no direct connection to the mains other than to charge the battery. Other batteries could be used, including a 4-S (four series cells, giving 14.8V) lithium-ion pack, but running time won't be too wonderful if the amp is driven to significant power levels. In contrast, a car battery (or 12V SLA battery) can be expected to give a reasonable listening time, depending on the capacity of the battery of course. A battery of not less than around 7AH (amp-hours) is suggested, and this should provide at least 10 hours of listening at modest levels.
As for the plank of wood to mount the amplifier on - needless to say that rare and exotic species of timber will have much better sound. It's even better if the tree is extinct or close to it. Coating the timber with French polish (preferably applied with a selected piece of yak leather rather than a cotton cloth) will improve the sound even more. 
It's hopefully obvious that I'm being silly - however the general principle is fine for experimentation and some people may even decide that they like the idea of a battery powered amplifier. There are no plans to design a PCB for this, but I will do so if there's enough interest. So, if the idea of a battery powered amp that's pretty much a PCB and a small heatsink (and little else) appeals, then feel free to let me know.
The main point of this article is to show that an amplifier of this type is easily built using relatively inexpensive ICs, and that paying a scam artist well over AU$1,200 is just silly. Why would anyone pay that for a rather ordinary amplifier just because it's presented with a vast amount of hype and BS? I don't know the answer, but I'm more than happy to mess with his plan because it's totally dishonest. For what it's worth, a long time reader alerted me to this particular piece of idiocy, and this article is my response.
There are quite a few power amp ICs available, but ideally the amp selected will be designed to give very good performance, and a reasonable output power with the limited voltage available. We certainly don't want to use output capacitors, so the amp has to use a BTL (bridge tied load) output stage. It doesn't help us that many of the ICs that used to be common are no longer available, having been replaced by Class-D in many cases.
Being Class-D isn't a reason to disqualify an IC of course, but many people are wary of low cost Class-D designs. They are more complex than an equivalent Class-AB IC, and also need output inductors and caps to filter out the switching frequency. Most are also surface mount, which makes it hard for many constructors. Accordingly, the first design shown uses the TDA7375 power amplifier. This is a highly refined IC, and it has features that most competitive devices lack.
Figure 1 - Circuit For The TDA7375 Power Amp
The input capacitors can be either 1µF polyester or any value from 2.2µF to 10µF electrolytic (shown as 'alternate'). My preference would be for the electrolytic, because it will be smaller and cheaper but will not compromise performance. The same applies to Figure 2, but isn't shown. With 1µF as shown, expect the amps to have a -3dB frequency of between 7Hz and 15Hz.
The IC is set up in such a way that it can be used as four independent amplifiers, or each pair of amps can be operated in bridge mode. This eliminates the need for output capacitors and allows higher power output than can be obtained with a single amplifier. The available power from a 12V battery isn't great though - even with bridged operation. For an actual voltage of 12V (rather than 14.4V as usually quoted), the absolute maximum power from a bridged amplifier is just over 17.5 watts into a 4 ohm load. Forget the silly numbers you often see in the datasheets - they allow for up to 10% distortion and also use other strange and mysterious formulae to produce numbers that look impressive, but are completely disconnected from reality.
The TDA7375 is designed so that it doesn't require external bootstrap capacitors, nor does it need Zobel networks (aka Boucherot cells). These typically employ a 10 ohm resistor and 100nF capacitor to ensure stability with reactive loads. The IC includes extensive protection circuits, and it will shut down if it overheats, or if an output is shorted to supply or ground. It will allegedly survive reverse polarity, but that's not something I'd rely on.
The pin marked 'DIAG' is a diagnostics pin, and the LED will show clipping (very brief flashes) or a fault condition with steady illumination. Both LEDs should be high brightness types because the available current is limited. Power is turned on/off using the standby (ST-BY) pin. When the voltage at this pin is close to zero, the amp is claimed to draw less than 100µA, so it won't discharge your battery. The pin marked 'SVR' is the half supply reference voltage, and 'SVR' stands for 'ripple voltage rejection' (in case you were wondering).
Figure 2 - Circuit For The Alternative TDA7297 Power Amp
The TDA7297 is similar to the TDA7375, but does not include the facility to operate each channel independently, and it's a dedicated BTL amplifier. It also lacks the diagnostic port, but has both standby and mute inputs. Ideally, the mute will be delayed for long enough to ensure that the input caps will charge before the mute is released, and this minimises turn-on thump. Somewhat remarkably, most of the pins have the same (or similar) functions, and the unused pins are not connected internally.
Because of the similarity of the two ICs, it would not be difficult to have a PCB layout that can accept either. This provides some much needed security of the design, because these ICs often have a fairly short lifetime in the market. It's not uncommon to find perfectly good power amp ICs that are now obsolete for no apparent reason.
The gain of these amps is fixed internally. The TDA7375 provides a gain of 26dB (x 20), and the TDA7297 has 32dB (x 40) gain. For full output, this translates to an input level of 212mV RMS for the TDA7375 and 106mV for the TDA7297. I have no idea why they are different, but those figures are as provided in the datasheets. In reality, the maximum input level will be a bit less than indicated because no car power amp ICs can swing their outputs to the full 0-12V available. Some are better than others in this respect, but in general expect that the maximum output swing per amplifier to be no more than ±5.5V from the 6V centre voltage. With a BTL amp, the total swing is therefore ±11V, or about 7.78V RMS. This implies a maximum output power of a little over 15W into 4 ohms at the onset of clipping. Output power in a car is higher because the battery voltage will be closer to 13.8V (although power is often given for a battery voltage of 14.4V which is not sustained in reality).
| Be aware that the ICs used do not provide any facility to shut down if the battery voltage falls below around 10.5V, the minimum allowable discharge voltage for a 12V lead-acid battery. That means that you need to ensure that the battery is charged regularly, and lead-acid batteries must never be kept in a discharged state for more than a few hours. A high quality 3-stage charger is the ideal for lead-acid, as it ensures the optimum charge cycle and keeps the battery in the best possible condition. Battery maintenance adds a certain ritual to the listening session. Never, ever, charge unsealed lead-acid batteries indoors. If they are overcharged, the battery will vent hydrogen and oxygen, and the mixture is highly explosive. |
Unfortunately, nearly all IC power amps need a PCB, because they have many closely spaced pins that refuse to line up with Veroboard or similar. The IC must have a heatsink, and it needs to be a reasonable size and have good air flow. The IC will shut down if it gets too hot, and that simply stops your listening session until it cools and restarts. As always, the cooler the IC runs the better, but you don't need to go overboard.
I haven't put one of these amps together (because I don't have a PCB), but the schematics shown are adapted directly from the datasheets, and they should both work as expected. Will they transport you to Nirvana? Almost certainly not, unless you are easily convinced by a bit of hocus-pocus. Naturally, you will never use either of the amps with a power supply that isn't a battery, but only partly because the current drain will be higher than your average 12V supply can deliver. The main reason (of course) is that the 'magic' will go away. 
One thing that does require some attention is the heatsink. While a very small one might look 'nice' (to some, although I disagree), it needs to be big enough to keep the IC's temperature within reasonable limits. A sure sign that the heatsink is too small is that the amp shuts itself down after it's been running for a while. Expect to dissipate up to 30W peak, although the average is likely to be no more than 10W or so. This means that the heatsink should have a thermal resistance of no more than 2-3°C/W. A larger heatsink will do no harm of course.
Whether or not you decide to use this project as an un-boxed board mounted on a piece of selected timber is entirely up to you. There are no actual (as opposed to imagined) benefits to this approach, but I suppose it could be seen as a talking point if nothing else. There will be some small issues with mounting the connectors (input, output and power), and there is no good reason to have them all mounted on the amplifier PCB. There are plenty of good reasons not to mount the connectors on the board - especially the speaker outputs.
The 'dubious characters' referred to in the introduction make some noise about the "length of the signal path", but this is obviously drivel, since both the source and speakers are required to have connecting cables. The length of the PCB tracks is immaterial, and only creates a system that's unwieldy and ugly if the connectors are all on the board. Don't forget to wave your anti-vibration magic wand over the finished amplifier, lest the vibration boogey-man comes around to ruin your simple pleasures.
Whether this is (or could become) a genuine project is a moot point. The ICs are undoubtedly fairly capable, and a car battery (or preferably a sealed lead-acid - SLA) battery disconnects the amp from the mains, but again, this is rarely a genuine problem. Added to the battery itself, you need a charger - another expense. If there is sufficient interest I will design a PCB for the amp that can use either of the ICs shown, but it's rather doubtful that this will happen.
Main Index
Projects Index