DIY ignoramus in need of help, opinions, improvements: TDA7560

[Formula350]

(I apologize for the length, I just want to try and give as much info about me and my situation/goal as I can so you all won't have to ask too many more questions)

Get this out of the way first-
Datasheet: TDA7560A - STMicroelectronics - 4 x 45 W quad bridge car radio amplifier

Alright so I've been into computers for a number of years, but never really got into it enough to pay mind to all the itty bitty IC parts, just motherboards, varrying chipsets, processors, video card etc etc etc... Recently I had a PSU die and since I didn't have the cash to buy one I thought I'd see if I could fix it... Long story short, no Soldered a few caps in backwards (reverse polarity) and they 'sploded literally in my face Don't worry, no harm came to any hardware during the testing of my epic fail, though sadly that PSU did kill (at least) 2 motherboards.

BUT that's not the point! Point is I've dove face first into this whole world, and as of only a week ago pretty much didn't know beans! The only reason I got into it was helping a friend research his X-Fi OpAmp (SMD) upgrade, and before then (again a week ago) I didn't know what that was This research turned up headphone amps, which lead to the 8pin DIP OpAmps that I noticed looked like the chip in my own little '90s KOSS speaker amp for walkman tape decks (I swear CMoy got the inspiration from as the design is so similar lol). I was going to port it over to a breadboard since it's PCB was deteriorating, but that ended in an epic fail as well :\ I redid it all last night with my slightly better understanding of things but it too failed

More on topic, I tore apart a Kenwood car DVD/Amp/DSP box-thing that I had no use for... and it had this TDA7560A (among 4 small JRC4565s and using all Elna caps, which sadly are 90% sub 50uF). I figured since I still have two 30x25 breadboards, an abundance of components from the various outdated (yet lightly used) PSUs, and a bunch of what I understand to be sort of decent Rubycon ZL caps from this ASUS built HP motherboard (even yielded an OS-CON heh), I figured I should be able to make this happen! While I would LOVE to make a HP amp with some of the OpAmps, their SMD nature and my lack of skill or soldering iron for that fine of work means I'll have to hold off on it.

So that brings me to the project and what help I need...

As I mentioned I'd like to build an amp. It'll be driving my 4 spare Logitech Z-560 sat speakers, but as of now no sub so it is of no concern. I'd just like non-headphone ability for audio until I get everything situated in my room (recently moved) and cleaned up enough to see where I sit. IF possible or worth doing (ie provide better sound), maybe give it the ability for me to plug in my intraaural earbuds; Shure SE210 (PDF for better specs). I understand that will be insanely overkill, but figured I'd ask. I did salvage this APA2308 if it is of use, and I'd be willing to try soldering leads onto it The hardest part is my budget; virtually nill. I am HOPING I have enough salvaged parts to get me through this, with the only likely required parts being resistors.

Amp specific items. This will be for my computer, so I'll be connecting to it through 3.5mm jacks. I have what is needed for that, but figured I'd mention it for the later "help" part. Seeing as this is orignally for a car, it means an operating voltage of not just DC, but 11 to at least 14.4v. Datasheet doesn't say a minimum but as I'm part grease monkey as well I would be surprised if it couldn't handle at least 11v, which doesn't matter really. Reason why is what I'd like to do; My computer's PSU is 1200W, nothing I have barely uses 1/2 of that, SO I have lots of spare power I can utilize... I'd like to power it from that using a molex connector (which provides 5v, 12v and dual grounds for anyone not familiar), which I again have covered but felt it worth mentioning. Heatsink is also not an issue, I have a bigger one than that it came with and worse case scenerio I could adapter a CPU heatsink and fan

IF anyone is still left at this point, I applaud your patience! This is where they will come in handy... The help I need is 2 fold really. I plan to use the "Standard test and application circuit" block diagram since it would probably suffice for my situation, but I'm hoping some of you will have a some suggestions where I can make improvements. I'll be happy to provide a brief list of inventory if needed, but if you think early 2000 era motherboards and power supplies you'll get an idea of components I have (not everything is crap like Teapo, lots are Rubycon or Nichicon).

The second bit is the actual block diagram :\ While the one in the PDF I have a much easier time of following, it sadly doesn't take much to confuse me I mean if a "Typical Circuit" diagram for a TDA2822M confused me enough to result in my setup not working, means I need it dumbed down for the mean time. Ideally a slightly more messy, but straight forward diagram. That being a pic of the chip that is exactly how it's pinout looks, instead of a rectangle with the pin numbers all out of order heh I'm pretty darn sure that if I can at least have one dumbed down enough and make it work, I'll be able to understand the circuit itself better and follow a typical diagram. As I said, I dove face first into this, I basically faceplanted on the ground and it isn't pretty

It's not all bad news though, I am a quick study and if anyone has the patience with me I think I can take my training wheels off before too long Also I have more recently been getting a lot of enjoyment out of researching stuff and learning. I've referenced out every little opamp looking chip I desoldered (unsoldered?) and read more datasheets in the past few days than most would be able to stand haha

So if anyone has any Qs or by some chance needs more info, I'll be happy to provide what I can!

Thanks!!
-Clint

 

[pacificblue]

There is actually little space for improvement, because all the important parts are inside the case. All you can do is experiment with the size and quality of the capacitors and the quality of the power supply. Computer PSUs are not always the best choice, because they tend to produce audible noises. If that is the case for your PSU remains to be tested.

With regards to the diagram, does the PCB layout in the datasheet not fulfill your requirements?

 

[Formula350]

There is actually little space for improvement, because all the important parts are inside the case. All you can do is experiment with the size and quality of the capacitors and the quality of the power supply. Computer PSUs are not always the best choice, because they tend to produce audible noises. If that is the case for your PSU remains to be tested.

With regards to the diagram, does the PCB layout in the datasheet not fulfill your requirements?

Fair enough Though to be honest, I would think a car would product far far more EMI than a computer power supply. You have the unfiltered alternator that revs up and down with the engine, and back in the day the ignition system would even cause bad AM radio interference (a whine that increased/decreased with the speed of the engine), which you can still get the kind that do but is rare. Though given that I've salvaged parts out of a ton of things, I would be surprised if I couldn't build a PSU for an amp, headphone or otherwise. Though I did plan on putting a ferrite choke on the power and ground inputs. Which brings up something I was wondering about: How much of a difference would putting chokes on the audio line inputs? I have three small ones that are perfect, but sadly they wrap around a ferrite core and not through a ferrite "doughnut". So I don't know how well they'd actually work :\ I did also plan on wiring up the power and ground with Networking (Cat5e) cable, which as odd as that may seem, should actually prove quite helpful. Reason being is the cables consist of 4 twisted-pairs of solid core wire, and as I'm sure I don't have to tell anyone here, that twisted wires help keep out EMI, especially if the negative/ground is wrapped with each wire. I had made a test audio cable to see how well it would work using that (for my headphones) and ran RMAA to test it against the standard cable that came with my Shure earbuds. And as to not get off topic, I'll forego the mention of the audible audio properties of that cable and just mention the relative part. My "DIY" cable vs the Shure resulted in the a fair amount less THD and noise (I mean not "OMG WOW", but more than what would be considered 'margin of error'). Also I had read an article months ago about building your own power wire for receivers and amps, where you run 2 ground wires and twist them around the Positive and Neutral power wires, which creates a magnetic field, blocking any EMI. So my theory is then that using twisted pair will help keep out some noise when using a computer PSU. Mind you it is pretty much an uneducated guess, as I don't truly know what twisted wires keep from happening, since it could very well be that any noise already in the line will not be 'eliminated' by twisting :\

Anyways it's not that the PCB layout that is my issue, it is understanding a block diagram (schematic if that is a better term), or rather being able to follow exactly where things need to go.

Example (even if ST's Datasheet example is easy and doesn't really confuse me): TDA2822M's block diagram shows the chip laid out like this-

2
7-------1
8-------.

6-------3
5-------.
4

Where as the physical chip layout (viewed from the bottom) is:

8-------1
7-------2
6-------3
5-------4

So when I'm looking at my breadboard, I can't as easily visualize the way the pins need to be connection in relation to the diagram. I've always been more of a hands-on and visual learner. Teach me something either laid out "plainly" or in person, and I'll pick up on it easier than having to study something and to learn how to do it. Might seem a bit "backwards", but maybe my brain is just wired that way lol Given that I've failed twice at trying to get this TDA2822M, I clearly don't know what I've done wrong, even if I say that the diagram is easy to read I tried rearranging the diagram in my head, but I think that might have been part of the reason why it failed, I thought I had it configured right when obviously not. That or I did and my not understanding of all the symbols correctly lead to a component being placed wrong. I've tried to research it and only now came up with something of more help, which apparently -| |- is non-polarized cap and -[] |- is polarized. So as I assumed that the "[]" meant the "white" side of the cap, it might have been the cause of it not working due to putting caps in backwards

I kind of screwed the TDA7560A up too much to be useful now anyway. The diagram I was looking at was for a non-A suffix model, and OF COURSE that meant different pinout on the A model And since my breadboard doesn't have the holes close enough to fit all the pins, I had bent the 4 pins for the input of channel 3 and 4, as well as the 6 output pins and 1 power pin for them. When I found out I was looking at the wrong datasheet and checked the rightit ended up being shifted one pin. So it meant having to bend 3 or 4 pins, get them aligned so they'll fit into the breadboard, and by that time one of the "TAB" pins broke off (NO clue what the TAB function is). On top of that I had thought ahead and since I wasn't going to make solder traces but use (what I call...) "jump wires" to go from the chip, to a spot on the PCB that was out of the way. Except with the different pinout of the A model meant all those wires were in the incorrect place

So now I'm on to using this LM358P and abandoning this 7560 project for the time being. I know that the 358 is a horrible chip for audio, but I figure that if I can at least get ONE flipping circuit to work, I might understand a bit better and feel up for trying a more difficult chip. IF I can get this 358 to work, my next plan is to use an LA6510 or BA10324A (again likely a poor choice but not my point right now haha). Even though I don't understand the LA6510's sample circuit at ALL We'll see I guess...

Oh and in regards to what you said about the caps and so forth. I follow what you mean, but I just didn't know if the capacity that the sample circuit lists is a safe amount and so using a larger size would be more efficient or yield more desireable audio.

I really do appreciate the help, and sorry for rambling again . I'm still open to help with this chip though, as I do plan at SOME point to come back to it.

​

 

[pacificblue]

a car would product far far more EMI than a computer power supply.

Yes, but in a different frequency range that is easy to filter with the capacitors you already have around the amp. Just try it, you may be lucky. Sometimes you get whining noises when you use SMPS, sometimes they are dead quiet.

How much of a difference would putting chokes on the audio line inputs?

Depends on its value and on the value of the other components that are present. There is always the risk that such a coil forms an oscillator with one of the adjacent capacitors. And inductive components in the signal path are usually not desirable.

I did also plan on wiring up the power and ground with Networking (Cat5e) cable

Several people report that they use Cat5 cable for the power supply with good results. As long as you make sure that the wire gauge is sufficient for the currents involved there is nothing wrong with that.

I tried rearranging the diagram in my head, but I think that might have been part of the reason why it failed

Probably. It helps to do that on paper. Draw the chip the way you see it and make the connections. It is easier to follow the wires with your finger or a pen on paper. After a while you will see that most diagrams become easier to read, when the numbers are swapped. Diagrams are usually drawn so that the signal flows from left to right and that voltage potentials are from top to bottom. The lower the voltage, the further is is down. Then you arrange the pins in a way that crossing lines are avoided as much as possible.

apparently -| |- is non-polarized cap and -[] |- is polarized. So as I assumed that the "[]" meant the "white" side of the cap

The solid black rectangle or line is the negative side, the white rectangle is the positive. The white side of the cap is the negative. Usually the - is printed on it as well. Turn them around on your PCB.

I just didn't know if the capacity that the sample circuit lists is a safe amount

It does. Depending on the speakers you drive it can be useful to increase the values of the input capacitors (C1-C4). That will give you deeper (not louder) bass.
Increasing C6 increases ripple rejection and at the same time the on/off delay.
Increasing C7 imrpoves power supply low frequency noise filtering and helps to maintain the supply voltage more stable under load.
C8 filter high-frequent noise. If you have trouble with oscillations it can sometimes help to decrease this capacitor, but you will probably not need to touch this.
C9 and C10 influence the on/off delay.

 

[Formula350]

Yes, but in a different frequency range that is easy to filter with the capacitors you already have around the amp. Just try it, you may be lucky. Sometimes you get whining noises when you use SMPS, sometimes they are dead quiet.


Depends on its value and on the value of the other components that are present. There is always the risk that such a coil forms an oscillator with one of the adjacent capacitors. And inductive components in the signal path are usually not desirable.


Several people report that they use Cat5 cable for the power supply with good results. As long as you make sure that the wire gauge is sufficient for the currents involved there is nothing wrong with that.


Probably. It helps to do that on paper. Draw the chip the way you see it and make the connections. It is easier to follow the wires with your finger or a pen on paper. After a while you will see that most diagrams become easier to read, when the numbers are swapped. Diagrams are usually drawn so that the signal flows from left to right and that voltage potentials are from top to bottom. The lower the voltage, the further is is down. Then you arrange the pins in a way that crossing lines are avoided as much as possible.


The solid black rectangle or line is the negative side, the white rectangle is the positive. The white side of the cap is the negative. Usually the - is printed on it as well. Turn them around on your PCB.


It does. Depending on the speakers you drive it can be useful to increase the values of the input capacitors (C1-C4). That will give you deeper (not louder) bass.
Increasing C6 increases ripple rejection and at the same time the on/off delay.
Increasing C7 imrpoves power supply low frequency noise filtering and helps to maintain the supply voltage more stable under load.
C8 filter high-frequent noise. If you have trouble with oscillations it can sometimes help to decrease this capacitor, but you will probably not need to touch this.
C9 and C10 influence the on/off delay.

I appreciate the info No matter how small (or trivial/common knowledge) to some, it's all a big help to me since I literally have little-to-no clue what I'm doing

I'm still putting this chip on help though, until I can at least get a first amp built, no matter how good or horrible the chip I end up using. I just want to get a first-run out of the way, so I can at least get my foot in the door. If I keep failing though, I might have to hold off for awhile or just simply stop trying heh

Thanks again!