Lm3886/tda1557

Hi guys,

I have got a 4 ohms speaker and a couple of power amplifiers. The first power amplifier is based on LM3886 capable of working at +-28V, and the second one is based on TDA1557 capable of working at 14.4V. When connected to a 4 ohms load the LM3886 will deliver a power of say 68W when powered by +-28V, and the TDA1557 when connected to the same load will deliver 22W at 14.4V supply to the said load.
Now the main question I have got is that I want to know how much of loudness or SPL difference I will get at the output of the said 4 ohms speaker when the speaker is connected either to the LM3886 at +-28V supply or the TDA1557 at 14.4V supply respective plz? Is the difference notable? Can I use the TDA1557 with the said speaker yet expecting a small reduction in the sound SPL? The main benefits of using the latter chip are using a transformer with lower secondary voltage and much less heating generated compared to the LM3886 when working at 68W.

Thanks for any input

The difference between 68W and 22W may seem like a lot but it is just a 4.9dB change in power (and sound pressure) which is noticeable but not a huge difference, since it is generally considered that, due to the non-linear response of the ear, a 3dB change is barely noticeable and 10dB is perceived to be about twice as loud.

crutschow said:

The difference between 68W and 22W may seem like a lot but it is just a 4.9dB change in power (and sound pressure) which is noticeable but not a huge difference, since it is generally considered that, due to the non-linear response of the ear, a 3dB change is barely noticeable and 10dB is perceived to be about twice as loud.

Click to expand...

Hi crutschow and thanks for your input.

Why did you calculate the SPL by 10log x? It should be 20log x I think.

The TDA1557 is an obsolete stereo car radio amplifier. Its output into 4 ohms is 22 Whats at a horrible 10% clipping distortion.
Its output into 4 ohms when clipping a little (0.5% distortion) is 17 Watts per channel.

Since the LM3886 is a single amplifier then the two channels in a TDA1557 playing at clipping will sound almost as loud as one LM3886 amplifier but the distortions are different.
The LM3886 is rated at 68W into 4 ohms with only 0.1% distortion and is only 0.03% at 60W into 4 ohms.

dr.power said:

Hi crutschow and thanks for your input.

Why did you calculate the SPL by 10log x? It should be 20log x I think.

Click to expand...

Look here.

audioguru said:

The TDA1557 is an obsolete stereo car radio amplifier. Its output into 4 ohms is 22 Whats at a horrible 10% clipping distortion.
Its output into 4 ohms when clipping a little (0.5% distortion) is 17 Watts per channel.

Since the LM3886 is a single amplifier then the two channels in a TDA1557 playing at clipping will sound almost as loud as one LM3886 amplifier but the distortions are different.
The LM3886 is rated at 68W into 4 ohms with only 0.1% distortion and is only 0.03% at 60W into 4 ohms.

Click to expand...

Hi guru,

Thanks for your input.

crutschow said:

Look here.

Click to expand...

Hi and thanks crutschow,
So you have calaculted the electrical/acoustical power of the amplifiers,

Can you guys tell me please how to calculte the SPL of an speaker while having the acoustical/electrical power of the connected amplifier? For instance is it easy to have an approach/relationship of the ouput SPL by knowing the power of an amplifier, for instance we know the power of the LM3886 is 68W at 4ohms when connected to a supply of say -+28V, So can we make a guess or calculte the ouput SPL?

Thanks a lot

Why guess about the sensitivity of a speaker? It is listed on its datasheet.
A cheap little speaker will probably not be loud but a quality big speaker will probably be extremely loud with the same amount of power from an amplifier.

But frequently a cheap little speaker has the gap from its voice coil to its magnet that is too small and it resonates loudly at one or two frequencies so it sounds awful but is very loud.

dr.power said:

Hi crutschow and thanks for your input.

Why did you calculate the SPL by 10log x? It should be 20log x I think.

Click to expand...

10log x for power, 20log x for voltage or current.