Some TDA7379 Amp questions

Fergutor

Member
2014-12-06 7:14 pm
I have a TDA7379 (now called STA540, apparently is the same chip, at least same specs) lying around and I want to use it.
I traced an Edifier C200 which is an excelent sounding system, at least with its factory speakers. Sounds very good, doesn't pick up any interference (and I live next to a f*****g radio antenna, has very low noise, it's very powerful, etc. The only problem it has is that it requires a lot of power from the source to properly drive it (in one review, from which I took some images showed here, the reviewer uses a preamp in front of it to counter such "problem").
But one more important problem for me now, is that it uses an 8-bit processor (EM78P447SAMJ) and an audio processor control tone (TDA7303), things that I will not use. So I wnted to copy at least some parts to make sure I get some of that good design. Then I will ask how to attach a bypass-able baxandall tone control.
But, I traced now from some photos I took some time ago (and few I found on internet), and I can't see some values (don't have here the system I took photos from, and mine is not open: one has to break the enclousure to open it and of course I don't want that).
Also another problem, it has inductors and no idea the value and where I live, chokes, inductors, ferrite cores, certain transformers and coils are almost impossible to get, so I'll ask how to do them myself. I will use a low quality 96Watts notebook universal charger to power this (I know...I know, but the C200 uses a switching PS too anyway...)

So what values do you think are the ones missing here? The ones that has values but with "?" are values I can almost read from the photos but I'm far from sure. The last caps are from a zobel network I think, and I will use initially some old Grundig 4ohm speakers, but then I will use others too with different values (originals aren't specified but they have a 6.5" 4 Ohms midwoofer and a 1" 6 Ohms silk dome tweeter, and testing with a multimeter gives me around 5 Ohms resistance). Also what modifications should I make given I will not use the stand by, mute, and diagnostic features? In the datasheet the standby is connected to something, what it is?
OUTPUT-Stage01.jpg

"J3" goes to headphones exit, "J12" to the front panel bard where the processor, display and controls are, C135/136 are connected to TDA7303 outputs I guess (can't see the traces there as they are under the chip).

This is L103:
inductorPS-7809.jpg
What value could it be? How do I do it myself?

Also, not in the schematic there's a sort of coil made with the cables from the TDA7379 to the zobel network and speaker output terminal.
OutputCoils01.jpg
00002285.jpg
Outputcoil02.JPG
Is it a coil? What's the purpose? How can I get away with that if I don't find a similar ferrite core (if that's what the cables are wrappled around on)?

Then the input stage, which are two actually (PC, RCA: 400mV ± 50mV / AUX, 3.5mm: 350mV ± 50mV, 10k input impedance, according to specs), so I added the auxiliar values. The difference seems negligible in use, both sound almost the same volume. This is an official schematic I found of it. As probably it's designed to work with the TDA7303, and thinking about it's low sensitivity (high power needed in input or risk low volume with cell phones and PC onboard soundcards) and given I will probably add a Baxandall tone control, what modifications should I make?
c200_61AUX.jpg

As you can see in the red circle, there's a coil there which is this:
InputCoils01.jpg
InputCoils02.jpg

I guess it's a HF filter, right? What value could it be? How do I do it if I find a similar ferrite core and if I find different ferrite core?

Other questions I have are:
-My universal charger can provide different voltages 12, 15, 16, 18, 19, 20, 22, 24v (with supposed enough amperage) given the C200 uses 15v (specs are 2x25Watts) and the TDA7379 (and STA540) can handle up to 22v for 2x 34W at 8ohms, 18v for 2x 38W at 4ohms and 18v for 2 x 20W at 8ohms, what voltage should I use? And what other filtering sould I apply?
-Do active baxandall tone controls invert the signal?
-Those yellow axial capacitors that can be seen in the photos, can I replace them for usual ceramic discs?
-Given I want to achieve good sound qulity, what other modifications should I make?

Thanks in advance.
 
Last edited:

Fergutor

Member
2014-12-06 7:14 pm
You see that type of filter if the designer knows that there is a problem or the designer finds it likely that there may be a problem. If no problem is foreseen, you don't add a quite large and relative expensive component. It is the first time I see it in this place. Perhaps it helps against ground loops?

And what about radio signals? I live next to a big radio antenna that mess with all I do, and this amplifier is inside a plastic case and it's completely inmune to that and to, apparently, anything else here. On the other hand I don't know if it would be a problem otherwise...
 
The designer clearly was very keen on keeping RF out and used the medieval fort approach. 1 nF - 820R - 680 pF makes for quite a lot of input capacitance. Apparently they had conditions such as yours in mind.

I am not exactly sure why the input stage has a -6 dB gain (not to mention input impedance of only around 3.3k), the TDA7379 only has 26 dB of gain in BTL mode. Perhaps they have some input gain active in the TDA7303 or didn't want to risk clipping its input with CD level signals. That chip's datasheet is a bit mystifying - it shows several graphs with a volume setting of up to +20 dB when the thing only has a +11.25 dB maximum input gain and no output gain. 's a pretty neat chip for micro systems and the like otherwise, with output noise levels that are easily hi-fi worthy and distortion that isn't quite so. Apparently it's a TDA7318 (now obsolete, dated 1999) with fewer inputs. Yeah, I did think the datasheet and performance had quite the '90s feel to it.
 

Fergutor

Member
2014-12-06 7:14 pm
The designer clearly was very keen on keeping RF out and used the medieval fort approach. 1 nF - 820R - 680 pF makes for quite a lot of input capacitance. Apparently they had conditions such as yours in mind.

I am not exactly sure why the input stage has a -6 dB gain (not to mention input impedance of only around 3.3k), the TDA7379 only has 26 dB of gain in BTL mode. Perhaps they have some input gain active in the TDA7303 or didn't want to risk clipping its input with CD level signals. That chip's datasheet is a bit mystifying - it shows several graphs with a volume setting of up to +20 dB when the thing only has a +11.25 dB maximum input gain and no output gain. 's a pretty neat chip for micro systems and the like otherwise, with output noise levels that are easily hi-fi worthy and distortion that isn't quite so. Apparently it's a TDA7318 (now obsolete, dated 1999) with fewer inputs. Yeah, I did think the datasheet and performance had quite the '90s feel to it.

Thanks for all the useful info.
So what do you think? Should I implement the same in my project? The common mode choke only (and how) or all until the choke (will this work with non inverting stage with single PS or dual ps)?
Yesterday I was looking at car amplifier shcematics (I'm interested in the crossover/filter section) and saw that they had what I think they call "differential balanced inputs" that apparently serve a similar purpose if I undertood that correctly. Should I better go with that simpler approach? Will it work for radio frequencies?
7520aInpuOut.jpg
("LOW INPUT R-CH" of course)