LM, TDA Chip Amps is **** Comparing with STK

how bass lost
 

Attachments

  • 20231106_204253.jpg
    20231106_204253.jpg
    368.3 KB · Views: 185
Give him a break, his command over English is not that good, please don't make fun
As for his post, he probably likes big bassy subs , brick walled music ,over driven beyond clipping level bass.
Can't blame this post Y2K generation for their listening taste
Not sure what us and europeans have to do with his messed up chip amp sub. Completely irrelevant. Not to mention political post after yours. Ignore button!
 
  • Like
Reactions: NanoFarad
But here one thing i should mention that, power amplifiers are not the same. Final outcome depends on the total system & it's execution. When i compared 3886 & 7294 using same speaker, TDA sounded soft(thin) at the low end compared to LM but that doesn't mean TDA is weak at BASS. With proper, dedicated spkrs 7294 is able to produce excellent bass. Many subwoofer system using 7294 at their power amplifier section.
 
The 10 Ohm output resistor is irrelevant at bass if the 0.7 uH inductor has a low dc resistance.
My suspicion is that the chip amplifiers are flat and low distortion. The STK is generating low frequency distortion that is heard as "more bass". Our ears are a bit insensitive down there and a bit of 2nd and 3rd harmonic is a common trick in bookshelf speakers.
 
The STK is generating low frequency distortion that is heard as "more bass".
What's the basis for this statement?
I built an amp with the Sanyo STK4048XII which has guaranteed THD specs of 0.005% max.
A spec like that is not going to produce audiable distrortion at any frequency.
It did measure flat and didn't have the pronounced bass sound you seem to be mentioning.
In fact, I still have it and use it often, some 35 years later.
I do know there were many versions in the STK series and the 4048XII was near the top of the line, so I can't really speak for the others.
 
If you pass a generator, you will see that there is nothing wrong with low frequencies using chip amps. The problem is in music that goes wrong. I have experience with LM3886, and TDA7293-4, as well as Chinese lm1875/TDA2050.
The problem is the PCB that is wrong. If you look the output and supply pins you see that the output is in sandwich with the supplies for the good reason that the thin leads of the chip cannot pass individualy high ac currents but they can If the magnetic fields are canceled. Those who design the PCB neglect to apply this fact.
A second reason is the positive feedback that the input applies to the ground. This you can suppress by using a virtual ground made of 2 resistors 2.2k from the supplies and use it as input ground. The input signal then enters the amp with a DC blocking capacitor.
 
  • Like
Reactions: availlyrics
The plot shows the response is flat to 10 Hz.
How it that a loss of bass?
Did you compare all the chips at the same gain settings?
As you may know, volume or loudness greatly affects our perception of bass frequencies, as in the Fletcher-Munson curves.
https://en.wikipedia.org/wiki/Equal-loudness_contour

tested input cap value and bs cap value

input 47uf
bs 100uf

simulator showing 10hz range, but lm3886 and tda7294 not produce bass frequency
 
My main gripe with chip amps from the times when they were the hottest fashion was the quality of the bass. Perhaps some highly sympathetic speakers would produce satisfying bass with the little critters, but none of mine at the time (Sonus Faber and ATC).

To make the bass a bit more palatable Jeff Rowland used a whole lot in parallel and perhaps this worked for him, but i have my doubts.

Cannot understand what is so controversial about someone not liking the bass of a particular device. Or any other subjective quality.
 
I suspect the answer to your problem 😉 one of them anyway would be taken care of in TI's design guide in the data sheet. An incorrect heat sink for instance is likely to cause problems. Values used in a 20hz to 20khz +/- 0.25db design example 40w into 4ohms seem to be rather different to yours.

Significant output at 20Hz is likely to need a rather large speaker and a correctly designed cabinet. On the car side if you like that sort of sound the usual answer is twice the power levels going to the woofer via a separate amp but you seem to be after a full range system.

Smoothing caps on the supply. They need to take account of the current load. There are calculators about on the web that give a decent idea of the ripple that will occur under max load or what ever load you choose. The transformer needs to be adequately rated as well. The data on those should allow you to work out temperature rise and the effects of regulation. The worst load is when sine wave testing,
 
  • Like
Reactions: availlyrics