I wanted to see what the maximum power output before clipping is on a few chip amps are, so I set up a test. Using a regulated 9 volt supply, non inductive 4 and 8 Ohm load resistors and a 200Hz sine wave source, I set up a test.
To find the highest clean output before clipping, I used an oscilloscope on the output to monitor the waveform. I increased the input signal until clipping began and backed it off just to the point of no clipping then took the RMS voltage reading across the load resistor.
The results:
LM386: This chip's weak drive current inhibits its ability to drive 4 Ohms loads. 8 Ohm loads are okay, but far from the best.
TDA2050: Although this IC is designed for much higher supply voltages, it's internal bias turns on at a very low voltage and it works fine for a low voltage amplifier (tested at +4.5/-4.5v). It drives 4 Ohms loads higher than the LM386 but is far from the "Swing King". Because it is designed it operate at much higher Vs, there is no need to optimize it's output swing like the low voltage ICs need so I included it here mainly for interest.
TDA7231: This middle aged (1980's tech) IC has improved output swing into 8 Ohms and much improved into 4 Ohms as compared to the LM386.
TDA1517: This is a later generation chip requires minimum parts and delivers a high output swing without need for a bootstrap capacitor. (one channel driven)
TDA2003: This middle aged IC was designed for maximum output swing for automotive use. It is about the same as the TDA1517.
TDA7276A: This later generation IC delivers the highest voltage swing than any of the others. It is the "Swing King" so far.
Vs-1.2v: this is the supply voltage minus the Vsat voltage of the upper and lower output transistors. I used .6 volts per transistor. The chart shows the maximum rms sine wave power (clean power) possible under this condition. With 8Ohm loads, many chips were close or tied it, but at 4 Ohm loads, for whatever reason, the ICs fell short.
To find the highest clean output before clipping, I used an oscilloscope on the output to monitor the waveform. I increased the input signal until clipping began and backed it off just to the point of no clipping then took the RMS voltage reading across the load resistor.
The results:
LM386: This chip's weak drive current inhibits its ability to drive 4 Ohms loads. 8 Ohm loads are okay, but far from the best.
TDA2050: Although this IC is designed for much higher supply voltages, it's internal bias turns on at a very low voltage and it works fine for a low voltage amplifier (tested at +4.5/-4.5v). It drives 4 Ohms loads higher than the LM386 but is far from the "Swing King". Because it is designed it operate at much higher Vs, there is no need to optimize it's output swing like the low voltage ICs need so I included it here mainly for interest.
TDA7231: This middle aged (1980's tech) IC has improved output swing into 8 Ohms and much improved into 4 Ohms as compared to the LM386.
TDA1517: This is a later generation chip requires minimum parts and delivers a high output swing without need for a bootstrap capacitor. (one channel driven)
TDA2003: This middle aged IC was designed for maximum output swing for automotive use. It is about the same as the TDA1517.
TDA7276A: This later generation IC delivers the highest voltage swing than any of the others. It is the "Swing King" so far.
Vs-1.2v: this is the supply voltage minus the Vsat voltage of the upper and lower output transistors. I used .6 volts per transistor. The chart shows the maximum rms sine wave power (clean power) possible under this condition. With 8Ohm loads, many chips were close or tied it, but at 4 Ohm loads, for whatever reason, the ICs fell short.
An externally hosted image should be here but it was not working when we last tested it.
I happen to have both of these ICs and can test. I suspect the 2822M will be similar than the TDA7231A (mono) and 2824 (stereo) as they seem to be similar amplifier design in different packages with the 2822m derated due to the limited ability to dissipate waste heat.
Thanks. I'd appreciate it if you'd test it in bridged mode - it packs quite a punch, too bad it gets quite a bit of hiss due to the ridiculous amount of gain when bridged.
Except for the fact that it has WAY too much hiss. Cant get rid of it on my end. Plus the stupid IC picks up every little piece of electrical noise in the ambient area. light switches, phone transmissions, etc...
Now this is an old thread...
The TDA2822 has very high gain of 39db, so noise (hiss) will be an issue. This can make it susceptible to receiving outside electrical noise, but that should manageable with proper shielding as with any amplifier.
Adding a resistor in series with the 100uf cap on pins 5 and 8 will increase the feedback and reduce the gain (with a side benefit of reducing distortion). You will have to experiment with the resistor value as the internal values are not given. Loop stability can be an issue at lower gain but I'd think you'd okay.
Last edited:
here is a way to reduce the gain from 40dB to 20dB, it shows the KA2209 but works as well with a TDA2822 here:
Mini Audio Amplifier Circuit using IC KA2209 [] Diagram Guide
Mini Audio Amplifier Circuit using IC KA2209 [] Diagram Guide
You don't need the gain. Even a modern source, such as an iPod, outputs 900 mV RMS. CD players and DACs generally go for 2 V RMS. A buffer or an amp with 2-4x gain (6-12 dB) is generally plenty to drive a pair of headphones. You do need a low noise floor as headphones tend to be rather sensitive. I'd look at the opamp+BUF634 or opamp+LME49600 (if you can still find it) schematics out there.
Tom
Tom
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.