Why did you think the tsx stock amp IC was a TDA7382? You could be right, especially since ST's selection of class d amps is quite limited. But the high freq distortion seen was unexpected for class ab. The distortion was seen with a non-inductive load and unloaded. Perhaps there was some inductance needed to limit high freq current during testing to stay within the gain-bandwidth product of the amp...

what are your thoughts?

 

I came across the cross reference during a replacement part search on a parts web site. It looked correct but I didn’t dig any deeper. I replace a lot of these modules in audio and guitar amps and I used them to build a 30 watt car amp with one back in the day when you couldn’t just go to the car audio store and buy one.

As far as the hi freq component in your output, I didn’t see any in my brief test a few weeks ago. I wasn’t looking that hard and I was mostly interested in the undistorted power output and the internal xover point. My scope is good to 150mhz so if it was there I should have seen it. By the way, the audio interconnects pick up a 1.2mhz clock from something in the car. I was able to lock my scope on it. This is a good reason to use a balanced line and amp.

Question, do you see the high freq noise with an input connected to the amp and with the cover on? If your audio source is not balanced it could easily introduce the noise or cause and oscillation to set up. It is unlikely (but not impossible) that the designer would layout an amp and have it oscillate at hi frequencies this would generate a lot of extra heat and shorten the device MTBF. I have run across an amp or two that needed some mods to quiet them down and prevent oscillation, it’s rare but it happens. These amps were all high power and high gain.

A class D amp would require a low pass filter and an A/D converter for each channel.
You could also search the other chip manufactures for similar chips.

jeff

 

maybe bandpass filters on the input stage would be helpful...

 

My freq response testing was not perfect and wish I had backtracked the signal to ensure it was only the amp IC creating high F distortion. I believe i did and the input was clean, but i can't say with 100% confidence. Testing with a balanced source would have also helped to prevent further confusion.

The amp was tested on my lab bench without the heatsink/shield on. Common mode issues could have played a role in the distortion. Mercman, you're certainly right about the pluses of a balanced input. Car audio should use nothing but balanced. The audio designer did well in that department and would have likely addressed oscillations like what i saw. Why I saw them... i'm not sure.

but I don't think i'll pursue testing the stock amp much further. I've learned enough to believe its not optimal for my desires. An amp with a flexible xover and adjustable gain per channel is what i'd like even if its just potentiometers added to a similar circuit.

class d's... the low pass filters are what I thought the designer screwed up on . a/d converters should not be needed though. check ti's description of the analog input of a class d amp.
http://focus.ti.com/lit/an/sloa031/sloa031.pdf

 

hey mercman...im interesting in getting a harness from u...

 

oops..interested