Building An Audio Amplifier

[Carel_Integrated]

Hi all, I'm trying to build an amplifier. I got an old car radio, so I removed the AMP from it. It's a TA8205AH, manufactured by Toshiba. Even though this is a 2-Channel 18W amp, it sure outpust far more than just the rated 18W & it sure can handle more bass than my BoomBox 150w amp... So, with only this chip & a few resistors & capacitors, I want to build an AMP. I'm actually done with it, I'm just having trouble with configuring the "Feedback condenser". If anyone would want to help me out, before I continue, you'll most probably need this datasheet: http://www.datasheetcatalog.org/datasheet/toshiba/1145.pdf
Ok, the amp is working, but as mentioned, I haven't configured the Feedback condenser yet. In the datasheet, there are 2x of them, they are marked C2 in the circuit diagram. Toshiba recommends 47uF. The formula to calculate C2 is: 1/(2pie * fL * Rf). What the heck is fL & Rf? I've tested the AMP with a lot of capacitors & the moment I attach them, the bass immediately increases & makes the audio sound more WoW, but then obviously I cannot set the volume too high, unless I reduce the bass. Does anyone care to explain to me what the heck is the "Feedback condenser" & what's the purpose of it? And what is "determination of low cut
off frequency"? I'm afraid to use the recommended 47uF {The Audio sounds WoW, but then the AMP cuts off if the bass is too high}. Please check the datasheet, it'll help a lot.

 

[Nigel Goodwin]

Even though this is a 2-Channel 18W amp, it sure outpust far more than just the rated 18W.

No, there's no way it can output more than it's design, the power supply can't provide any more. Personally, assuming it's a bridged amplifier, I would rate it at 16W, not 18W - and even at 16W distortion is rather high.

As for capacitor values, see what the original radio used.

 

[condor36zw]

On one of the tables f is given as 1KhZ, which in this case is the low pass cuttoff frequency, i.e. the upper limit of your Bass.

How they get Rf fL, I have no idea. RL I would understand, as this is given as 4Ω, the load resistance. Maybe a misprint, which would give you:-

C2=1/(2pi X RL X f)
=1/6.283 X 4 X 1000)
= .000039
=39uF....closest 47uf, but electrolytics have crap tolerances anyway.

I dunno

 

[flat5]

Try a few caps, like 22uf, 16, 10uf, and see what you like.
Do it with the speaker you plan to use in it's enclosure.
Play music you plan to listen to.
In other words, use your ears to select the cap.

 

[audioguru]

The datasheet claims an output of 18 Whats per channel when the supply voltage is too high at 14.4V and the volume is turned up too high so it has 10% awful clipping distortion. It shows it beginning to clip when its output is only 10 Watts per channel into 4 ohms when its supply is a reasonable 13.2V.

 

[Carel_Integrated]

Ok, for the capacitor, I've used 47uF. It gives the most Bass, but then of course I cannot set the volume too high. It was a choice between:
1.) High Bass, Low Volume
2.) Lower Bass, High Volume

So, I picked option 1. This amp really has a very, very high sensitivity & emmits a lot of noise & ripple from the speakers... I'll have to use a High Definition Audio device for the input, because this amp will amplify the noise coming from the source too. When I tested it on my PC, I had to set the volume of the PC to 5% & it was already kicking hard... The amp is my 1st attempt & I'm looking forward to the next. I really would've liked this amp, but it's just very noisy. If I disconnect the source, then the noise are very quiet, but if I connect the source, I can hear a lot of ssssshhhhhh & zzzzzzzzz, except if I turn the volume up, then it seems to make less noise for some reason. I think my biggest problem here is the input, because seriously, setting the PCs volume to 5% is very low. {All audio devices emmit a sssshhhhhhh sound, even HDMI devices. It's very low, but it's still there. On High Definition Audio devices, it's not possible to hear the noise, but this amp for sure detects noise & then amplifies it}. I think the only thing I need to do is use some dumping resistors at the input, because even at 1% I can clearly hear audio playing from this amp {If I use my Hi-Fi, it's signal processor doesn't even detect audio at 1%}.

1.) What the heck is the difference between Class A, B, C & D amplifiers & which of them are the best?
2.) What Class amp would this be?
3.) Does anyone perhaps know of some other amp I can build that outputs 2-Channel audio between 15 ~ 45W @ 4 ~ 6ohms, maybe a circuit diagram?

 

[audioguru]

A class-A amplifier has very high current in its output transistors all the time so it wastes a lot of power making a lot of heat.
A class-B amplifier has no current in its output transistors at idle but it has severe crossover distortion at low signal levels.
A class-C amplifier is cutoff at idle and usually drives pusles into a tuned circuit that smooths the pulses into a sine-wave.
A class-D amplifier switches its output transistors on and off with pulse-width-modulation at a high frequency to reduce heating.

Most audio amplifiers including this one operate in class-AB where the output transistors have a low bias current to avoid crossover distortion but not enough current for them to get hot.

The Rf value for the amplifier circuit determines the sensitivity. If you don't have a resistor Rf then the gain is 50dB which is way too high (313 times).

A TDA7240A single amplifier IC has a little more output power and much lower distortion. Its gain is usually 100 times. A pcb design is in its datasheet. Use two for stereo.

 

[Carel_Integrated]

Ok, it seems that the noise have been eliminated completely now. I've put a 1K-ohm resistor between {IN1 & NF1} and {IN2 & NF2}. And then I've connected a 10K-ohm resistor between {IN1 & the Audio Source} and {IN2 & the Audio Source}. This have eliminated the zzzzzz on long wires between the Audio Source & the AMP's input, and it eliminated the hum & shhhhhhh coming from the Source. I'ts quiet now & it seems to be an excelent AMP. I will definitly be using this thing, for use with a PC, it's WAY better than my BoomBox 150W 2-Channel amp. Hey, I've opened the BoomBox amp, just to see what IC's they have been using & they haves used ST Microelectronics chips {I can't remember what's the part no}. Those 2x chips are smaller than Toshiba's TA8205 & they cannot provide tha same quality of audio. In my opinion, the BoomBox 150W amp is a piece of you know what. The damn thing isn't even capable of knocking 50W speakers to their max, it's definitly an OVERRATED amp. If I would rate this amp I built myself TA8205 18W 2-Channel vs. BoomBox 150W 2-Channel, then I would consider the BoomBox I bought as Junk... In most cases custom built items are crap, but in this case I would rate this car Radio AMP as far better as the one I bought. It delivers more Bass, Clearer Audio {After a lot of struggling} & it doesn't get as hot as the BoomBox. Hey - At least I have an AMP that's not "Made In China" {which in my opinion is the worst manufacturer of anything you can think of in terms of quality & performance}, Toshiba & Philips is from Japan {which in my opinion is just simply the best in the world, in terms of Efficiency, Failure Rates, Performance & definitly Quality}.

 

[unclejed613]

Why a 16W amp is louder than a 150W amp:
**broken link removed**

i don't know for sure i would even call it a 16W amp, since the slope of the THD curve is already almost vertical by the time you get to 15W. it looks more like the actual clipping begins at 10W. if you want a good sounding chip amp with more power (you will have to build a suitable power supply for it) that acts more like a true "power op amp", look at this one:
**broken link removed**

it's a much better amplifier. the gain setting is a LOT easier to calculate. the data sheet is here:
https://www.electro-tech-online.com/custompdfs/2009/12/LM3886.pdf

a class C amp isn't used for audio (unless you count the class C portion of a class G amp which is a class AB with class C switches connecting a higher voltage set of power supplies)

 

[audioguru]

Philips is not Japanese. It is Dutch, from Holland in Europe.

They used to make a class-H car amplifier that had a built-in supply voltage doubler. It switched from "low power" 15 Watts into 4 ohms to "high power" 50 Watts into 4 ohms when the signal needed the high power so that the amplifier stayed cooler. Its distortion was very low.
It was the TDA1562Q.

 

[Carel_Integrated]

Philips not Japanese - Oops, I didn't know... But I knew for sure it's not China {Their failure rate is very low & they provide only High Quality Chips}...

I've just opened the BoomBox & discovered that they used 2x TDA2005 chips, manufatured by ST Microelectronics - "Made In China"... Datasheet: https://www.electro-tech-online.com/custompdfs/2009/12/tda2005.pdf

Good CRAP, this AMP I bought as a 150W 2-Channel. It was advertised as capable of 300W Mono Block.... And the chips they used are only 10W + 10W. The Chips are each bridged. So i'ts a real cheapo amp & it for sure sucks.... My Hi-Fi uses the STK405-110 {http://www.audiolabga.com/pdf/STK405-110.pdf}. The STK405-110 provides excelent audio, Clear, Loud & High Bass.

Overratings SUCK... Crap Manufacturers, bad habits. PMPO is nothing else than Overrating. In other words, multiply the products's capability with +- 50. Shucks, so this AMP is overrated by almost 800%. I'll never buy one again.

 

[audioguru]

Most high quality European, American and Japanese audio equipment manufacturers have their own factories in China. They produce very high quality products. No-Name-Brand Chinese products are mostly garbage.

STK hybrid amplifiers have fairly poor performance and are made by Sanyo of Japan. Low distortion is rated when the output is only 5W. Their max output power is rated at only 10% distortion which sounds awful.

A TDA2005 bridged amplifier has an output of 15W at clipping into 4 ohms when the supply is 14.4V or 15W at clipping into 3.2 ohms when the supply is 13.2V.

 

[Carel_Integrated]

Ok, I'll attempt to build another AMP {not today, next year}. I think I'll be using 2x LM3886 chips, if I'm able to find it here... If I don't find it, I now know the drill for the TA82XX chips, so then I'll try to find a TA8254 or equivalent, it really impressed me with it's low noise, but it was very difficult to configure... There's just 1 thing I don't like about it at all: I cannot connect the Audio Source to the same power supply as the TA82XX series, because the speakers (-) doesn't use the ground. So, if I connect this amp to a computer's PSU & try to use the PC's sound, then it's absolutely muted, unless if I use a battery. Hopefully the LM3886 wouldn't do this, because it uses Output pin as (+) & the Ground as (-).

1.) I guess I can build it {LM3886} with the exact Schematic as seen in the datasheet, under "Single Supply Application Circuit"???
2.) Would I benefit if I "upgrade" from STK405-110 to LM3886? {building a custom AMP, the STK405-110 can still safely operate at 125°C, i'ts an 11-Year old IC}.
3.) I'm currently using a Switching Power Supply, using 2x BJT's, to provide 12 ~ 17V & +- 20A. I'm using a Switching Power Supply instead of a Transformer, Transformer's tend to produce a humming sound. I know Switching Power Supplys provide cleaner VCC, has less Ripple & are very energy efficient, but it's failure rate is higher, it's sensitive to Spikes, Lightning & Heat {the operating temperature has a direct influence on the Safe Operating Area}. I hope that using a Switching PSU wouldn't be a problem, I only have access to a 12V Terminal?
4.) Are MOSFET AMPs better than these ones?

 

[unclejed613]

to "upgrade" from the STK packs would probably require a discrete amp. there are higher powered STK chips, but their distortion and noise ratings are about the same, and without a major overhaul of the power supply (power transformer of higher voltage AND current ratings, higher current bridge rectifier, higher voltage rated filter caps), you won't get more power out of it anyway. i made up a chart for the audio department where i work so techs could "ballpark" output power, based on rail voltages. it's quite handy to have. i'll see if i can post a copy of it here. back to the original subject, you can get much better distortion and noise specs out of a discrete amp, but it takes up more board space and takes more skill to build (and a lot of skill to figure out what's wrong with it if something goes wrong). there are many designs for discrete amps on the web, but usually it's just the schematic, so you have to do the board layout yourself, and with audio power amps that is something that MUST be done right. moving a feedback resistor a half an inch either way on the speaker line trace can put an otherwise fantastic distortion figure back into the terrible category just as having traces too long to a compensation cap can turn your amplifier into an AM radio transmitter (until the output transistors burn up, which tends to be between a few milliseconds to about 5 seconds).